WO2020221540A1

WO2020221540A1 - Reduction of nitrogen oxide in the exhaust gas flow of a firing system with an scr catalyst

Info

Publication number: WO2020221540A1
Application number: PCT/EP2020/059211
Authority: WO
Inventors: Roland Pohle; Alexandru POPESCU; Henning Schramm; Maik Springer
Original assignee: Siemens Aktiengesellschaft
Priority date: 2019-05-02
Filing date: 2020-04-01
Publication date: 2020-11-05

Abstract

The invention relates to an assembly for reducing nitrogen oxide in the exhaust gas flow (9) of a firing system (6), having: at least one SCR catalyst (1), at least one ammonia injection grid (2) which is arranged upstream of the SCR catalyst (1) in the exhaust gas flow direction and which is divided into multiple segments, throttle devices (3) which control the supply of ammonia to the ammonia injection grid (2), each of the segments being supplied with ammonia by a respective throttle device (3), an exhaust gas flue (7.1) which is arranged downstream of the SCR catalyst (1) in the exhaust gas flow direction, and a measuring device (8) which is designed to ascertain the ammonia concentration and the nitrogen oxide concentration of the exhaust gas flow (9) in the exhaust gas flue (7.1). The invention is characterized by: sensors (4.1) which are arranged downstream of the SCR catalyst (1) and are distributed over the exhaust gas cross-section and which are arranged and designed to ascertain the local distribution of the sum concentration of ammonia and nitrogen oxide, wherein for each of the segments, at least one spatially corresponding sensor (4.1) is provided, and a computing and control unit (5) which is designed and programmed to control the ammonia supply by means of the throttle devices (3) on the basis of the local distribution of the sum concentration while taking into consideration the ascertained ammonia concentration and the nitrogen oxide concentration in the exhaust gas flue (7.1) such that specified target values for the ammonia concentration and the nitrogen oxide concentration are reached. The invention also relates to a corresponding method.

Description

description

Reduction of nitrogen oxides in the exhaust gas flow of a combustion system with an SCR catalytic converter

Field of invention

The invention relates to an arrangement and a method for reducing nitrogen oxides in the exhaust gas stream of a furnace system with an SCR catalyst.

Background of the invention

SCR catalytic converters in power plants (coal, gas, oil) and other combustion plants require ammonia (NH ₃₎ to reduce nitrogen oxides. In the following, these systems are also referred to as combustion systems. Ammonia is injected into the exhaust gas flow upstream of the SCR catalytic converters via an ammonia injection grid (AIG). The dimension of the exhaust gas duct cross-section determines the division of the injection grille into several individual segments.

The reason for the division into several segments is that different NO _x concentrations can prevail at the respective positions in the flue gas duct, which then also require locally different amounts of ammonia for optimal operation of the combustion system. The local adaptation of the amount of ammonia injected to the locally existing NO _x concentration in the exhaust gas is particularly important when very high NO _x conversion rates and low NO _x or NH ₃ slip concentrations have to be achieved, as is increasingly required.

For this purpose, each of the segments of the AIG is individually supplied with the reducing agent ammonia via a line. In each of these lines there is a valve, which in most cases is designed as a simple manual control valve. In the course of commissioning the combustion system, the individual control valves are adjusted so that the required NO _x limit values and the NH ₃ slip values are achieved over the entire cross-section of the exhaust gas duct. After the optimal setting has been found, the control valves are fixed in the "optimal" setting.

The problem here is that an optimized manual setting of the valves is only carried out for a single load case (in most cases the full load case), but the uneven distribution of the NO _x concentration over the exhaust gas duct cross-section depends on the operating state (full load - part load - Minimum load) of the system. Another problem is that the non-uniform distribution of the NO _x concentration over the exhaust gas cross-section changes with increasing operating time, for example due to aging of the SCR catalytic converter.

Thus, a manual, one-off setting cannot guarantee the optimal operation of the SCR catalytic converter over the long term, as a result of which an increase in the emissions of NO _x and ammonia is to be expected over the operating time.

SCR (selective catalytic reduction) refers to a catalytic technology for the reduction of nitrogen oxides in exhaust gases from combustion plants, waste incineration plants, gas turbines, industrial plants and internal combustion engines. The chemical reaction on an SCR catalytic converter is selective, which means that preferably the

Nitrogen oxides (NO, NO _{2) are} reduced, while undesirable side reactions such as the oxidation of sulfur dioxide to sulfur trioxide are largely suppressed.

As an option to adjust the setting of the ammonia grid (ie the ammonia mass flows to the individual segments) to the operating conditions that have changed over time, the control valves are regularly readjusted at intervals of several years. However, this is not ideal for the following reasons: - The effort for readjustment is very high, since a so-called network measurement of the NO _x concentration must be carried out over the cross-section of the exhaust gas duct. In some cases a measuring grid is installed before and / or after the SCR catalytic converter for this purpose. In many cases, such a network measurement is not possible due to the structural features.

- As with commissioning, the setting of the control valves only ever applies to a restricted operating range (in the most common cases the full load case). Increasingly, however, compliance with low slip values is also required in other load cases (partial load, start-up) in which a significantly different NO _x concentration profile can develop over the boiler cross-section. Compliance with these low slip values (NO _x and NH ₃₎ is then not given for the other load cases.

A continuous measurement of the local NO _x concentration in SCR systems and a continuous readjustment of individual segments of the ammonia injector grid is currently not carried out for reasons of cost. The reason is that very expensive measuring devices are usually used to measure NO _x concentrations, which allow measurements to be taken at only a few points in the exhaust gas duct and / or in the exhaust stack (e.g. laser-based infrared measurement technology, chemiluminescence analyzers).

Constant network measurements cannot be carried out cost-effectively with it. A readjustment or regulation depending on operating conditions (load changes) that are in some cases rapidly changing is therefore not economically feasible with the measurement technology currently in use. In addition to the local NO _x concentration, the flow rate prevailing at the respective point naturally also has an influence on the required amount of ammonia to be injected. The amount of ammonia to be injected locally is ultimately the product of the local NO _x concentration and the local gas velocity. A so-called network measurement of the exhaust gas velocity over the exhaust gas duct cross-section is in most cases not feasible.

Summary of the invention

It is the object of the invention to provide a solution with de Ren reduced nitrogen oxides in the exhaust gas flow of a combustion system with an SCR catalytic converter under all load conditions who can.

The invention results from the features of the independent claims. The dependent claims relate to advantageous developments and refinements. Further features, possible applications and advantages of the invention emerge from the following description.

An automatic control of the control valves of the ammonia injector grid with low-cost sensors distributed over the boiler cross-section is proposed. The measurement of the NO _x concentration can be done easily, quickly and inexpensively with such low-cost sensors (such as those used in large numbers in the automotive sector) in contrast to conventional measurements as network measurements at a large number of positions in the exhaust duct .

Due to the continuous measurement of the NO _x concentration and the activation of the ammonia control valves as a control, the influence of the local flow velocity also has no effect. For example, if the local flow velocity is relatively high, the ammonia mass flow that is injected is correspondingly more diluted, but the measurement shows this and thus leads to a correspondingly higher NH ₃ injection until the concentration corresponds to the target concentration.

The low-cost sensors for NO _x measurement are those developed for monitoring the SCR systems in diesel engines Sensors into consideration. These are offered, for example, by the companies Bosch and Continental as automotive suppliers. In terms of their measurement properties, these sensors are in principle suitable for use in accordance with the invention. The measurement accuracy is typically specified as ± 10 ppm at 90 ppm NO _x . In certain applications, however, these sensors achieve a resolution of <1 ppm and detection limits of <1 ppm. However, a service life of typically only is used for motor vehicles

2000 h - 3000 h indicated. However, it is also known that a service life of> 10000 h can be achieved under other conditions of use, although a change in the detection properties must be taken into account. The following challenges arise with this procedure proposed by the applicant:

- The low-cost sensors used in the automotive sector generally measure NO _x and NH3 equally. They thus provide a sum signal of both concentrations. An independent determination and thus an adjustment of both concentrations is therefore not possible without further information (e.g. about the operating status of the system).

- The existing low-cost sensors are designed for a significantly shorter service life, which may not match the service life or availability of sensors required in the power plant sector.

- According to the manufacturer's specifications, the sensors have a limited absolute accuracy that lies within the range of the required emission limit values or is worse.

- In the course of their service life, the sensors have a drift, which leads to deviating measurement signals over time.

In addition, it is proposed that the control (sensors, control of the control valves) be integrated into a cloud (= cloud computing) in accordance with the approaches of Industry 4.0 / digitization. Through the use of machine learning and artificial intelligence, the challenges mentioned can be mastered:

1. At the chimney, the total NO _x and NH3 concentration is usually measured separately using high-precision analysis technology. This information should be combined with the measured values of the low-cost sensor grid and change the currently used injection logic to such an extent that the total NO _x and NH3 concentration at the chimney is optimally minimized and thus below the levels to be observed Limits falls. The following must be observed: a) The influence of individual segments on the total NO _x and NH3 concentration at the chimney can be determined from the time course of the NO _x / NH3 concentration of the low-cost sensor grid under different load conditions . b) By evaluating the temporal change in the current measured values of individual low-cost sensors, a trend for the future can be determined with constant NH3 injection. This trend can be used to change the local NH3 injection to such an extent that the total NO _x and NH3 concentration at the chimney is minimized. c) The spatial distribution of the measured values of the low-cost sensors can be resolved over time and also shows a spatial and temporal trend for different load conditions that change the spatial NO _x concentration. The spatial and temporal change in the low-cost sensor values is characteristic of the stationary and transient load state, so that an optimized spatial NH3 injection can be calculated from it. The latter is determined in such a way that the total NO _x and NH3 concentration in the chimney is minimized.

2. In addition, statements can be made about the aging of individual Kataly satorsegmente in terms of "predictive maintenance". 3. Local NH3 and NO _x concentrations can be determined separately from one another by evaluating large amounts of data.

4. A failure or a drift of individual sensors can be recognized by evaluating temporal progressions and taken into account in the regulation. In this way, the service life of the sensors can be used optimally and reliable operation of the system can be achieved even if individual sensors fail.

Not only the individual sensors of a combustion system, but all the sensors of several systems are connected to the cloud. This enables conclusions to be drawn from the data of others

Power plants. The joint evaluation of this data enables further conclusions to be drawn about the optimization potential of an individual system and a system fleet. The latter represents a considerable added value for the applicant as a system manufacturer, since it enables optimized solutions based on historical data to be offered, which were previously not possible due to the lack of available history and data.

With the connection of the low-cost sensors to a cloud and the application of machine learning and artificial intelligence, the use of low-cost sensors for monitoring and optimizing ammonia injection is reliably possible. This makes a network measurement economically viable and this is the only way to automatically control the control valves of the ammonia injector grid. This enables compliance with low emission values (NO _x and NH3) in the entire load range and during load changes in the power plant. The regulation is tuned continuously and with minimal effort.

In addition, there are the following advantages:

- Through appropriate "training" of the algorithms, future events such as failure of individual sensors or al- the SCR catalytic converter can be closed. Corresponding measures, such as changing the SCR catalytic converter, no longer have to be carried out regularly, but can be carried out as required.

- By using machine learning, some combustion systems equipped with the proposed technology (distributed sensors) can deduce the control of the NH ₃ grid of systems without this technology (e.g. via substitute measured values such as ambient temperature, air humidity, load).

- The failure of individual sensors can be caused by e.g. Measured values from other sensors in the grid or using data from sensors from other systems can be compensated.

- By minimizing the ammonia slip, considerable amounts of operating media (ammonia) can be saved compared to conventional operation of SCR catalysts.

The invention claims an arrangement for reducing nitrogen oxides in the exhaust gas flow of a combustion system, comprising: at least one SCR catalytic converter,

at least one ammonia injection grille arranged in front of the SCR catalytic converter in the direction of the exhaust gas flow and divided into several segments,

the ammonia supply to the ammonia inlet grille controlling throttle devices, each one of the Drosselvor devices one of the segments is supplied with ammonia, an exhaust gas chimney arranged in the exhaust gas flow direction after the SCR catalyst,

a measuring device which is designed to determine the ammonia concentration and the nitrogen oxide concentration of the exhaust gas flow separately in the exhaust gas stack,

several sensors arranged after the SCR catalytic converter and distributed over the exhaust gas cross-section, which sensors are arranged and designed to determine the local distribution of the total concentration of ammonia and nitrogen oxide, with for each of the segments has at least one spatially corresponding render sensor, and

a computing and control unit, which is designed and programmed, depending on the local distribution of the total concentration, including the determined ammonia concentration and the nitrogen oxide concentration in the exhaust gas chimney to control the ammonia supply through the throttle devices in such a way that predetermined setpoints for the ammonia concentration and the Nitric oxide concentration can be achieved.

The computing and control unit preferably has a neural network that has been trained from historical measured values and / or measured values from further combustion systems.

In a further development, the throttle device can have a valve or a flap. This ensures accurate injection of ammonia.

In a further development, the computing and control unit can be designed and programmed to determine the influence of the segments on the ammonia concentration and the nitrogen oxide concentration in the flue gas chimney from the temporal course of the total concentration of the sensors at different load states of the combustion system.

In a further embodiment, the computing and control unit can be designed and programmed to determine a trend for the future ammonia concentration and the future nitrogen oxide concentration with a constant ammonia injection from the change in the total concentration of one of the sensors over time.

In a further embodiment, the computing and control unit can be designed and programmed to change the ammonia injection into the segments based on the trend determined in such a way that the ammonia concentration and the nitrogen oxide concentration in the exhaust stack are minimized. In a further embodiment, the computing and control unit can be designed and programmed to determine a spatial and temporal trend from the spatial distribution and temporal resolution of the cumulative concentrations of the sensors for various load states of the combustion system that change the spatial ammonia concentration and from the spatial and temporal Trend to change the ammonia injection into the segments in such a way that the ammonia concentration and the nitrogen oxide concentration in the exhaust chimney are minimized.

In a further embodiment, the SCR catalytic converter can be formed from catalytic converter segments which are arranged spatially corresponding to the segments of the ammonia inlet grille, and the computing and control unit can be designed and programmed, a measure for the aging of a catalytic converter segment in the sense of from the total concentration preventive maintenance.

In a further development, the computing and control unit can be designed and programmed to approximately determine the local distribution of the ammonia concentration and the local nitrogen oxide concentration at the SCR catalytic converter from the sum concentrations, the ammonia concentrations and the nitrogen oxide concentrations in the exhaust stack of a large number of load states.

In a further embodiment, the computing and control unit can be designed and programmed to determine a failure or drift of individual sensors from the temporal progressions of the total concentrations and to take this into account when regulating the ammonia injection.

As a result, the service life of the sensors can be optimally used and reliable operation of the arrangement can be achieved even if individual sensors fail. In a further embodiment of the invention, the firing system can have a gas turbine.

A gas turbine is an internal combustion engine in which a fuel is burned to generate (mechanical) power.

In a further embodiment, the computing and control unit can be connected to a cloud.

Cloud or cloud computing (German computer cloud or data cloud) is an IT infrastructure that is made available via the Internet, for example. It usually includes storage space, computing power or application software as a service.

In a further embodiment, the ammonia induction grid, the SCR catalyst and the sensors can be arranged in a heat boiler.

The invention also claims an automated method for reducing nitrogen oxides in the exhaust gas flow of a furnace, with an arrangement which has:

at least one SCR catalytic converter,

a measuring device which is designed to determine the ammonia concentration and the nitrogen oxide concentration of the exhaust gas flow separately in the exhaust gas chimney, and

sensors arranged after the SCR catalytic converter and distributed over the exhaust gas cross-section, which are arranged and trained, the local distribution of the total concentration to be determined from ammonia and nitrogen oxide, with at least one spatially corresponding sensor being available for each of the segments,

depending on the local distribution of the Summenkon concentration taking into account the determined ammonia concentration and the nitrogen oxide concentration in the exhaust chimney, the ammonia supply through the throttle devices is controlled in such a way that predetermined setpoints for the ammonia concentration and the nitrogen oxide concentration are achieved.

Further features and advantages of the invention will become apparent from the following explanations of an exemplary embodiment with the aid of schematic drawings.

Brief description of the drawings

Show it:

1 block diagram of a furnace,

2 shows a grid-like arrangement of the sensors in a first embodiment,

3 shows a grid-like arrangement of the sensors in a second embodiment,

4 shows a block diagram of a furnace in a first embodiment,

5 shows a block diagram of a furnace in a second embodiment,

6 shows a block diagram of a combustion system in a third embodiment,

7 shows a block diagram of a furnace in a fourth embodiment, 8 shows a block diagram of a combustion system in a fifth embodiment,

9 shows a block diagram of a combustion system in a sixth embodiment and

10 shows a block diagram of a combustion system in a seventh embodiment.

Detailed description of the invention

Fig. 1 shows a block diagram of a combustion system, which indicates the general principle of operation. From the gas turbine 6, the exhaust gas flow 9 passes into the waste heat boiler sel 7. The exhaust gas flow 9 leaves the waste heat boiler 7 cleaned through the waste gas chimney 7.1. In the direction of the exhaust gas flow 9, an ammonia injector grid 2, an SCR catalyst 1 and a sensor grid 4 are arranged one behind the other in the waste heat boiler 7. The ammonia injector grille 2 is divided into several segments, with each segment being supplied via a throttle device 3 with ammonia for injection into the waste heat boiler 7.

The SCR catalytic converter 1 can also be divided into segments corresponding to the segments of the ammonia inlet grille 2. The sensor grid 4 has sensors 4.1 which are arranged spatially corresponding to the segments for measuring the total concentration of ammonia and nitrogen oxide. The sensors are preferably low-cost sensors, as described in detail above, and known from exhaust gas cleaning in motor vehicles. In the exhaust stack 7.1 there is a measuring device 8 for the separate and very precise determination of the ammonia concentration and the nitrogen oxide concentration in the exhaust gas stream 9.

The measurement signals from the sensors 4.1, the measurement signals from the measurement device 8 and the operating data from the gas turbine 6 are transmitted to the computing and control unit 5. From the arithmetic and control unit 5, the throttle devices 3 can be steered so that, depending on the load, less or more ammonia is spatially and selectively injected into the waste heat boiler 7. The regulation takes place in such a way that, depending on the local distribution of the total concentration, including the determined ammonia concentration and the nitrogen oxide concentration in the flue gas chimney 7.1, the ammonia supply through the throttle device 3 is controlled in such a way that predetermined setpoints for the ammonia concentration and the nitrogen oxide concentration are achieved

Fig. 2 shows a schematic plan view of a lattice-shaped arrangement of the sensors 4.1 in a first embodiment. The sensors 4.1 arranged on a sensor grid 4 determine the cumulative concentrations of ammonia and in a spatially resolved manner in a level downstream of the SCR catalytic converter 1 in the exhaust gas stream, not shown Nitric oxide. The measured values of the sensors 4.1 are fed to the computing and control unit 5 (not shown) via the sensor signal lines 4.2. It can have a neural network for evaluating the measured values and for controlling the throttle devices 3.

Fig. 3 shows a schematic plan view of a gitterför shaped arrangement of the sensors 4.1 in a second embodiment. In contrast to the first embodiment

2, several spatially distributed total concentrations are determined by one or more sensors 4.1. For this purpose, the sensors 4.1 are connected to pipelines 4.3, whereby the total concentration at the far end of the pipelines 4.3 can be determined. The sensors 4.1 are in turn connected to the computing and control unit 5 (not shown) via the sensor signal lines 4.2.

Fig. 4 to Fig. 10 show block diagrams of a combustion system in seven different embodiments. What all seven embodiments have in common is that in the firing system from a gas turbine 6, the exhaust gas flow 9 is a waste heat boiler 7 is supplied. The cleaned exhaust gas flow 9 leaves the waste heat boiler 7 through the exhaust gas chimney 7.1. A measuring device 8 for the separate and very precise determination of the nitrogen oxide concentration and the ammonia concentration is arranged in the exhaust gas tank 7.1.

In the waste heat boiler 7 are located in the exhaust gas flow direction in front of the exhaust chimney 7.1, the divided ammonia induction grid 2, the SCR catalyst 1 and the sensor grid 4 with the sensors 4.1 spatially distributed on the sensor grid 4 (not shown), which have a total concentration of ammonia and nitrogen oxide can determine. The control of the injections in the ammonia injector grid 2 is carried out by throttle devices 3 corresponding to the segments, for example by valves or throttle flaps.

In the first exemplary embodiment according to FIG. 4, the signals from the measuring device 8 and the sensors 4.1 as well as to and from the throttle devices 3 are transmitted to and from the computing and control unit 5 of the firing system controller 12. The gas turbine 6 is also connected to the fire control system control 12. In the computing and control unit 5, as described under FIG. 1, the measured values are evaluated and the throttle devices 3 are controlled as a result.

In the second embodiment according to FIG. 5, the signals from the measuring device 8 and the sensors 4.1 as well as to and from the throttle devices 3 to or from the computing and control unit 5 of the firing system controller 12 are transmitted. The gas turbine engine 6 is also controlled by the Combustion system control 12 connected. In the computing and control unit 5, as described under FIG. 1, the measured values are evaluated and the throttle devices 3 are controlled as a result.

The furnace control system 12 is connected to a cloud 11. The connection of the combustion system control 12 in one ne cloud-based database provides long-term predictions from historical data and enables predictive maintenance with an SC catalyst management plan.

In the third exemplary embodiment according to FIG. 6, the signals are transmitted from the measuring device 8 and the sensors 4.1 to the edge device 10. The edge device calculates the gas concentrations from the sensor signals and transmits the gas concentrations to the combustion system control 12. The signals to and from the throttle devices 3 are transmitted to and from the computing and control unit 5 of the combustion system control 12. The gas turbine 6 is also connected to the combustion system controller 12 in terms of control technology. In the computing and control unit 5, as described in FIG. 1, the evaluation of the concentration values and the resulting control of the throttle devices 3 take place.

The fourth exemplary embodiment according to FIG. 7 is the same as the third exemplary embodiment according to FIG. 6. In addition, the firing system control 12 is connected to a cloud 11.

The connection of the combustion system control 12 to a cloud-based database supplies long-term predictions from historical data and allows predictive maintenance with an SCR catalytic converter management plan.

The fifth exemplary embodiment according to FIG. 8 is similar to the fourth exemplary embodiment according to FIG. 7. In contrast to the fourth exemplary embodiment, the computing and control unit 5 is embodied in the edge device 10.

The sixth exemplary embodiment according to FIG. 9 is similar to the fifth exemplary embodiment according to FIG. 8. In contrast to the fifth exemplary embodiment, the edge device 10 is connected to the cloud 11 and not the combustion system controller 11.

The seventh embodiment according to FIG. 10 is similar to the sixth embodiment according to FIG The sixth embodiment lacks a central firing system control 11.

Although the invention has been illustrated and described in more detail by the exemplary embodiments, the invention is not restricted by the disclosed examples and other variations can be derived therefrom by the person skilled in the art without departing from the scope of protection of the invention.

List of reference symbols

1 SCR catalytic converter

2 ammonia inlet grids

3 throttle device

4 sensor grids

4.1 sensor

4.2 Sensor signal line

4.3 Pipeline

5 computing and control unit

6 gas turbine

7 waste heat boiler

7.1 Exhaust chimney

8 measuring device

9 exhaust gas flow

10 Edge Device

11 cloud

12 Combustion plant control

Claims

1. Arrangement for reducing nitrogen oxides in the exhaust gas flow (9) of a combustion system (6), comprising:

at least one SCR catalytic converter (1),

at least one ammonia injector grille (2) arranged upstream of the SCR catalytic converter (1) in the exhaust gas flow direction and divided into several segments,

throttle devices (3) controlling the ammonia supply to the ammonia injection grid (2), one of the throttle devices (3) supplying one of the segments with ammonia,

an exhaust gas chimney (7.1) and arranged in the exhaust gas flow direction after the SCR catalytic converter (1)

a measuring device (8) which is designed to determine the ammonia concentration and the nitrogen oxide concentration of the exhaust gas flow (9) separately in the exhaust gas min (7.1), characterized by:

in the exhaust gas flow direction after the SCR catalytic converter (1) and distributed over the exhaust gas cross-section sensors (4.1), which are arranged and designed to determine the local distribution of the total concentration of ammonia and nitrogen oxide, with at least one spatially corresponding one for each of the segments Sensor (4.1) is present, and

a computing and control unit (5) which is designed and programmed to control the ammonia supply through the Drosselvor devices (3) depending on the local distribution of the cumulative concentration, including the determined ammonia concentration and the nitrogen oxide concentration in the exhaust gas chimney (7.1) in such a way that predetermined Target values for the ammonia concentration and the nitrogen oxide concentration should be achieved.

2. Arrangement according to claim 1,

characterized,

that the computing and control unit (5) has a neural network.

3. Arrangement according to claim 1 or 2,

characterized,

that the throttle device (3) has a valve or a throttle flap.

4. Arrangement according to one of the preceding claims,

characterized,

that the arithmetic and control unit (5) is designed and programmed to determine the influence of the segments on the ammonia concentration and the nitrogen oxide concentration in the flue gas chimney (7.1) from the temporal course of the total concentration of the sensors (4.1) at different load states of the combustion system (6) ) to determine.

5. Arrangement according to one of the preceding claims,

characterized,

that the computing and control unit (5) is designed and programmed to determine a trend for the future ammonia concentration and the future nitrogen oxide concentration with a constant ammonia injection from the change in the sum concentration of one of the sensors (4.1) over time.

6. Arrangement according to claim 5,

characterized,

that the computing and control unit (5) is designed and programmed to change the ammonia injection into the segments based on the determined trend in such a way that the ammonia concentration and the nitrogen oxide concentration in the exhaust gas min (7.1) are minimized.

7. Arrangement according to one of the preceding claims,

characterized,

that the computing and control unit (5) is designed and programmed, from the spatial distribution and temporal resolution of the cumulative concentrations of the sensors (4.1) a spatial and temporal trend for various load conditions of the fire that change the spatial ammonia concentration Erungsanlage (6) to determine and from the spatial and temporal trend to change the ammonia injection into the segments in such a way that the ammonia concentration and the nitrogen oxide concentration in the exhaust chimney (7.1) are minimized.

8. Arrangement according to one of the preceding claims,

characterized,

that the SCR catalytic converter (1) is formed from catalytic converter segments which are arranged spatially corresponding to the segments of the ammonia injector grid and that the computing and control unit (5) is designed and programmed, a measure of the aging of a catalytic converter segment from the total concentration in terms of preventive maintenance.

9. Arrangement according to one of the preceding claims,

characterized,

that the computing and control unit (5) is designed and programmed, the local distribution of the ammonia concentration and the local distribution of the nitrogen oxide concentration on the local distribution of the ammonia concentration and the local distribution of the nitrogen oxide concentration from the total concentrations, the ammonia concentration and the nitrogen oxide concentration in the exhaust gas min (7.1) SCR catalytic converter (1) to be determined approximately.

10. Arrangement according to one of the preceding claims,

characterized,

that the computing and control unit (5) is designed and programmed to determine a failure or drift of the sensors (4.1) from the temporal progressions of the sum concentrations and to take this into account when regulating the ammonia injection.

11 arrangement according to one of the preceding claims,

characterized,

that the furnace has a gas turbine (6).

12. Arrangement according to one of the preceding claims, characterized,

that the computing and control unit (5) is connected to a cloud.

13. Arrangement according to one of the preceding claims, characterized in that

that the ammonia injection grid (2), the SCR catalyst (1) and the sensors (4.1) are arranged in a waste heat boiler (7).

14. Automated process for the reduction of nitrogen oxides in the exhaust gas flow (9) of a combustion system (6), with an arrangement that has:

at least one SCR catalytic converter (1),

an exhaust gas chimney (7.1) arranged in the exhaust gas flow direction after the SCR catalytic converter (1),

a measuring device (8) which is designed to determine the ammonia concentration and the nitrogen oxide concentration of the exhaust gas flow (9) separately in the exhaust gas tank (7.1), and sensors (4.1) arranged after the SCR catalytic converter (1) and distributed over the exhaust gas cross-section ), which are arranged and designed to determine the local distribution of the total concentration of ammonia and nitrogen oxide, with at least one spatially corresponding sensor (4.1) being available for each of the segments,

characterized,

that depending on the local distribution of the Summenkon concentration taking into account the determined ammonia concentration and the determined nitrogen oxide concentration in the exhaust gas chimney (7.1), the ammonia supply through the throttle device (3) is controlled such that predetermined setpoints for the ammonia concentration and the nitrogen oxide concentration can be achieved.