EP2064490B1 - Method for characterizing the exhaust gas burn-off quality in combustion systems - Google Patents

Description

The invention relates to a method for characterizing the Abgasausbrandqualität in incineration plants according to the first claim.

One goal of technical combustion processes is to achieve as complete as possible an efficient exhaust gas burn-out. Efficient exhaust burnout is characterized by low levels of incomplete combustion products such as CO, hydrocarbons and particulate carbon (soot particles). Emission limit values for this are usually specified in relevant regulations. In Germany, for example, the 17th BImSchV (Federal Immission Control Ordinance) sets the limit values for carbon monoxide CO and hydrocarbons C _n H _m .

Fuels such as household waste, biomass or coal with fluctuating moisture contents are very inhomogeneous fuels. Due to their very heterogeneous composition, their calorific value varies greatly. When incinerating in technical firing systems, complex firing-power regulations with infrared detectors (IR camera, infrared camera) are therefore used today in the combustion chamber. The firing position of the solid fuel bed in grate firing is determined using the infrared radiation of the fuel bed with the help of an IR camera. The wavelength detected here (for example, 3.9 μm) is in a range in which combustion gases themselves have no emissivity. With the help of this information, the control of the grate kinematics and / or the individual primary gas flows, which flow through the fixed bed. As a result, a nearly complete Feststoffausbrand the slag can be achieved.

An exhaust gas which emerges unevenly burned from a combustion chamber, for example a fixed bed burnout zone, generally has locally high concentrations of incompletely burnt compounds, such as CO, hydrocarbons and soot. The gas flow emerging from the combustion bed shows a pronounced formation of strands with enormous local and temporal fluctuations in concentration the aforementioned incompletely burned compounds as well as the oxygen concentration. These strands extend through the Abgasausbrandzone in the first Strahlungszug. For a homogeneous mixing and thus a complete combustion of the exhaust gas, the available mixing time or even the mixing turbulence is often insufficient. An incomplete burnout of the exhaust gases is therefore encountered with an introduction of an oxygen-containing secondary gas in the Abgasausbrandzone. The total amount of this secondary gas is chosen so that behind the Abgasausbrandzone always a defined excess of oxygen (minimum oxygen concentration) is maintained. The minimum oxygen concentration is limited by the required minimum combustion temperatures after the exhaust gas burnout zone.

In DE 10 347 340 A1 discloses an apparatus for optimizing exhaust burnout in incinerators having a fixed bed burnout zone and an exhaust burnout zone. It comprises a plurality of controllable nozzles for introducing oxygen-containing secondary gas into an effective region in the exhaust gas burnout zone. The detection of the individual incompletely burned gas components (CO and hydrocarbons) in the effective range via a detection of the radiation intensity by means of infrared camera or other spectrally measuring device. The information thus obtained is converted into control commands for each of the controllable nozzles for the targeted introduction of secondary gas.

However, the device and the associated method are used for the non-selective detection of incompletely burned gaseous components in the exhaust gas. Both incompletely burned gases and solid components (eg soot) are recorded as a sum signal, whereby a weighting between individual components is not possible. In addition, it may happen that areas where no combustion activities take place at all due to a lack of combustion gases are also detected as incompletely burned waste gas areas (cross sensitivity emissivity of CO ₂ to H ₂ O). In the latter case, an injection of an oxygen-containing secondary gas would cause no afterburning, but only a dilution and cooling of the gases.

Proceeding from this, the object of the invention is to propose a method for the characterization of the flue gas combustion quality with respect to Rußausbrand in incinerators as the basis for optimizing Abgasausbrandes especially for a complete Rußausbrand even with transient combustion processes with a minimum of secondary gas.

The object is achieved by a method having the features of claim 1. Returned dependent claims indicate advantageous embodiments of the method.

To achieve the object, a method is proposed for characterizing the exhaust gas burnout quality of a combustion in incinerators with a gas burn zone, wherein the soot particles, d. H. Solid particles in the exhaust gas can be selectively detected.

An essential basic idea of the method involves the relationship that in a flow cross-section of the gas burnout zone low-combustion regions (preferably without soot formation), regions without combustion and regions with soot formation in the visible wavelength range are optically detectable. The soot-poor combustion areas always appear bright (high radiation intensity), while the areas without combustion (cold rust areas) and sooty areas always appear dark (low radiation intensity). The combustion areas darken increasingly with increasing amounts of soot, i. the radiation intensity decreases continuously with the soot content. In this case, the areas without combustion and the sooting areas characterized by a different dynamics in their temporal behavior, which can be detected by an evaluation, preferably an averaging or a comparison of several consecutive individual recordings.

A basic prerequisite for the method is at least one camera system with camera measuring in the visible wavelength range (about 400 to 1000 nm), for example a video camera, which is adapted to a gas burnout zone in such a way that it has a flow cross section as completely as possible. In contrast to detection systems for the infrared range or other non-visible wavelength ranges, such camera systems are available as sophisticated standard systems for different applications, comparatively inexpensive and also available in high quality and high resolution on the market.

The camera system is used to record the combustion in the flow cross section with a sequence of individual shots. The single shots are snapshots of the Abgasausbrands in the entire flow cross-section, with camera setting and image section between the individual shots are not changed. The image detail preferably corresponds to the flow cross section in the region of the exhaust gas burnout zone. This flow cross-section is subdivided into segments with a number of pixels for evaluation of the images (image processing). The evaluation essentially comprises an assignment of the segments to one of the aforementioned areas or to transition areas between two areas by means of the method steps described below.

At least two of the successive individual images are averaged (preferably pixel-by-pixel) in order to generate an average image from them. In this average image, soot-poor combustion areas are recognized by their intensity value (radiation intensity) being above an adjustable intensity threshold value. The intensity threshold value is determined manually or automatically relative to the maximum intensity in the acquired image (for example 50, 60 or 70% relative to the respective maximum value) or manually specified as the absolute value. A manually specified intensity threshold can be composed of previous empirical values and remains in successive measurements in favor of improved comparability of these measurements, e.g. preferably unchanged for system monitoring.

Subsequently, the localization of the transition regions takes place as the pixels which have been assigned to the russar-combustion region but which have at least one adjacent pixel which is not belongs to the russarmmen combustion area. Thereafter, these transition areas are assigned to the transition segments. In this case, each segment is evaluated as to whether it must be attributed to one or as a transition segment at least two of the aforementioned areas, ie to a transition area. A transition region exists when the determined intensity corresponds to the intensity threshold value or a transition occurs between a value less than the threshold value and a value greater than the threshold value. In a single image or an averaged image, the mostly linear transition areas can be highlighted as lines, eg in color (eg false color representation). A transitional segment is basically present if the intensity threshold is both exceeded and undershot. The assignment of segments to transition segments is usually carried out using adjustable Flächenanteilsschwellwerte for the area proportions of the above individual areas.

Subsequently, the transition segments are assigned to the participating areas. A transition is indicated by a brightness difference. This can be determined, for example, by brightness gradients or segment by segment by determining a contrast that is calculated by means of a cooccurrence matrix (see www.weblearn.hs-bremen.de/risse/AWI/textur/merkmale.htm).

Basically, transitions from a low-carbon combustion region to a non-combustion region are characterized by lower dynamics of motion than soot transitions, i. Transitions from low-carbon combustion areas to sooting areas. In the case of averaged images formed from a plurality of individual images, carbon black transitions are generally characterized by a low selectivity or a lower contrast, ie. they appear much more blurred than transitions from a low-carbon combustion area to an area without combustion (rust transitions), although this may not be the case with single shots.

An assignment of the transition segments to the fraction of carbon black transitions or rust transitions is preferably carried out by determining the contrast separately for each transitional segment. An assignment of the transition segments takes place with contrast values in comparison to a contrast threshold value. If the contrast value of a segment lies below the contrast threshold, if there is a soot transition segment, it lies above it, a rust transition.

The contrast values preferably relate to the light intensity (light dark contrast). Other contrasts, such as Color contrasts e.g. In connection with a color manipulation of the images are in principle also for the aforementioned classification, but may require a higher billing expenses and are therefore preferred for a timely characterization of Abgasausbrandqualität only in special cases.

Alternatively, as part of an assessment, it is conceivable to distinguish between carbon black transitions and rust transitions by comparing individual transition segments or another group of pixels of the transition regions from successive individual images. Larger and faster changes in the intensity values of a segment or a group of pixels from a plurality of consecutive individual images indicate increased dynamics in the transition region and thus in soot transitions.

If contiguous transition segments of a linear transition region are not uniformly evaluated, but mixed both in the fraction of carbon black transitions and in the roasting transitions, an optional weighting of the individual fractions takes place. One possible method step involves the recognition of contiguous transition segments of a fraction and of individual transition segments of a fraction which are surrounded by the respective fraction. In this case, with a significant overweight of transition segments of one of the fractions all transition segments of this fraction can be assigned. Individual segments of a fraction can also be assigned to the fractions of the neighboring segments via a neighborhood analysis. On the other hand, contiguous transition segments of a fraction are only assigned to the other fraction if they are considered as Possible faulty measurements represent a single event (plausibility check).

Following the abovementioned assignment of the transition segments, an iterative assignment of all segments in which the intensity of more than half of the pixels lies below the intensity threshold takes place to the sooting area or to the area without combustion by evaluation of neighborhood relationships to transition segments and already assigned segments. The assignment of these segments takes place individually in iterative method steps by assuming the affiliation of the respectively adjacent, already already identified segments or transition segments (neighborhood analysis). In case of a non-uniform affiliation of the already assigned adjacent segments, the segment is assigned to the area to which most of the adjacent segments have already been assigned. Each of the iteration steps is preferably carried out on the segments which adjoin, as far as possible, a fraction to the largest possible number of already assigned transition segments or already assigned segments.

Finally, as part of the characterization of the exhaust gas quality, the individual layers, the surface expansions and the intensity distributions of all identified regions are determined. From these parameters, control variables for measures to improve the quality of exhaust gas burnout, such as a calculated, preferably a spatially differentiated (preferably segment-wise or segment-wise), adapted to the local combustion state input of oxygen-containing gases (for example, secondary gas, sooting areas) or additional fuels (in areas without combustion) calculated.

If the method of generating control signals for a measure, with the aim of continuously improving soot burnout (eg targeted injection of oxygen-containing gas), must be based on the individual recordings in the context of implementing the process in real time, the determination of the parameters. Through a multiplicity of gas nozzles, afterburning in each segment can also be influenced individually by supplying an oxygen-containing gas.

• Fig.1 eine mittels einer CMOS-Kamera aufgenommenen Einzelaufnahme eines Querschnitts einer Gasausbrandzone,
• Fig.2 eine aus 20 aufeinander folgenden, innerhalb einer Sekunde erfassten Einzelaufnahmen wie Fig.1 gemittelte Aufnahme,
• Fig.3 den Ausschnitt gemäß Fig.2 , jedoch mit Übergangssegmenten zwischen rußarmen Verbrennungsbereich und rußenden Bereich (helle Umrandung) sowie zwischen rußarmen Verbrennungsbereich und Bereich ohne Verbrennung (dunkle Umrandung),
• Fig.4 den Ausschnitt gemäß Fig.2 und 3 nach über Nachbarschaftsbeziehung erfolgter iterativer Zuordnung der dunklen Segmente in kalte Rostbereiche (Bereich ohne Verbrennung, dunkle Umrandung) und rußende Bereiche (helle Umrandung) sowie
• Fig.5 ein Kennfeld mit Bereichen mit effizientem Abgasausbrand am Beispiel von Kohlenmonoxid CO als Funktion von Verbrennungstemperatur und Sauerstoffgehalt. Die Charakteristik der Rußkonzentrationen verhält sich ähnlich wie die von CO.

The invention will be explained in more detail with examples with reference to figures. Show it

• Fig.1 a single shot of a cross section of a gas burnout zone recorded by means of a CMOS camera,
• Fig.2 one of 20 consecutive, within a second captured individual shots like Fig.1 averaged recording,
• Figure 3 the clipping according to Fig.2 But without combustion transition segments between low-soot combustion area and sooting region (bright border) and between the low-soot combustion area and region (dark border),
• Figure 4 the clipping according to Fig.2 and 3 after iterative assignment of the dark segments into cold rust areas (area without combustion, dark border) and sooting areas (bright border) as well as through neighborhood relations
• Figure 5 a map with areas with efficient Abgasausbrand the example of carbon monoxide CO as a function of combustion temperature and oxygen content. The characteristic of the soot concentrations behaves similar to that of CO.

As part of the following experimental example, the Abgasausbrand a waste incineration was characterized with grate firing. The image section of the camera according to Fig.1 to 4 detects from above against the gas flow direction the Strahlungszugquerschnitt the Abgasausbrandzone between the combustion grate and a downstream secondary combustion chamber with secondary gas input option. As a camera in the visible wavelength range measuring camera, such as a CMOS camera is used.

Fig.1 shows a single shot of Abgasausbrands with bright low-combustion area 1 and one dark rust area 2 (area without combustion) and a strongly sooting combustion zone 3 (sooting area) with low radiation intensity. The transition regions between these regions show similar brightness gradients on this single image, which do not permit a clear assignment of the respectively adjacent dark region to the region without combustion or to the sooty region.

The aim of the invention is to automatically identify and classify these regions with low radiation intensity on the basis of several individual images, and whether they are areas with a high proportion of soot (sooting area) or cold rust areas. Preferably, this should be done in order to initiate targeted actions such as additional gas injection in real time.

Fig.2 shows one from 20, within a second consecutive single shots accordingly Fig.1 average recording. In this mean value image, in contrast to a single image (cf. Fig.1 ) the boundaries between combustion 1 (flame) and sooting area 3 very blurred, which is due to the high dynamics of soot particle movement in the flow field. Due to the low dynamics of the boundary between the cold rust area and the low-carbon combustion area compared to the soot, this is consequently still relatively sharp in the average image.

This difference in the characteristics of the transition areas between combustion and soot-free area is used in the further method to distinguish soot areas 3 from areas without combustion (cold grate areas 2) from one another. For this purpose, first a boundary between the high-intensity radiant low-carbon combustion region and the region without combustion or the sooting regions is determined on the basis of a relative threshold value of the radiation intensity and entered in the averaging image as transition line 4 (transition region) (cf. Figure 3 , gray line). The mean value image is subdivided into segments, and those segments are determined which cover as transition segments 5 the transition line 4 between strongly and weakly radiating regions.

For the mentioned transition segments 5 , a contrast analysis is then used to determine whether there is a segment with a boundary between the area without combustion (rust area) and low-carbon combustion area, a grate transition segment 6 ( FIG. Figure 3 , Black border segments) or a segment of a boundary between rußendem area and low-soot combustion area, a Rußübergangssegment 7 ( Figure 3 , White-edged segments) is.

In addition, the integral intensity of each of the transition segments is preferably checked with the intensity threshold value at a later point in time, and the transition segments above the threshold value are assigned to the soot-poor combustion regions. Alternatively, each transition segment may also be assigned to the low-carbon combustion region if the intensity of at least half of the pixels of this segment is above the intensity threshold.

Subsequently, it is iteratively determined for all other segments outside the transition segments, which span the areas with weak radiation intensity, by evaluating the respective neighborhood relationships to the transition segments, whether they belong to a cold rust area or to a strongly sooting area (cf. Figure 4 ). The dark areas of the image section of the mean value image are thus subdivided into grate segments 8 (including the grate transition segments 6) and soot segments 9 (including the soot transition segments 7). In the present embodiment, the areas were uniquely identified. A plausibility check was not required here.

Basically, household waste, but also biomass with fluctuating moisture contents, very inhomogeneous fuels (and consequent large fluctuations in the calorific value), which favor not only the low-combustion cold grate areas (areas without combustion), but also incomplete combustion (sooting areas). These fuel properties lead to a different igniting and burning behavior. In technical furnaces (eg grate firing, fluidized bed, Rotary tube), it comes, due to this fuel characteristic, to local inhomogeneities in solid burnout and the exhaust gas composition (exhaust strands) within the combustion chamber and in the field of Abgasausbrandzone. The position and intensity of these exhaust strands also have pronounced temporal and local fluctuations, wherein the sooty regions in principle have a considerably higher dynamics. .

By the invention, these inhomogeneities in the exhaust gas or fuel gas preferably in the aforementioned real-time in the combustion chamber / Abgasausbrandzone by measurement, ie optically detected and compensated by controlled targeted local oxygen-containing gas supply and / or effective mixing so that at high temperatures and sufficient oxygen supply (above about 5 vol.% dry oxygen in the raw gas, T> 850 ° C, cf. Figure 5 ) a virtually complete oxidation of the incompletely burned exhaust gas components in a short time is possible.

In the Figure 5 revealed concentration of carbon monoxide CO in the fuel gas are like the soot concentration an indicator of the burn-out. The characteristic of soot burnout is similar. A good Aüsbrand is characterized by low concentrations of CO, C _n H _m and soot. They are essentially dependent on the local oxygen supply and on the temperature in the region of the exhaust gas burnout zone. The measured values in Figure 5 each show significant increases in carbon monoxide and hydrocarbon at temperatures below 800 ° C (in Figure 5 left measured values) and with oxygen contents below 5 vol.% (in Figure 5 right measured values), while the measured values shown in the middle signal a satisfactory burnout (ideal: T> 850 ° C and O ₂ > 5 Vol.%). The latter values correspond, for example, to the temperature of at least 850 ° C. prescribed for waste incineration in accordance with 17. BImSchV within a residence time of more than 2 seconds after the last oxygen-containing air addition. These conditions must be maintained at all times and throughout the entire cross section of the exhaust gas burnout zone.

In particular, minimizing the concentration of soot particles by efficient burnout has a very important importance in waste incineration. Soot particles deposit on the boiler surface together with chloride-containing fly ash or are separated during dedusting (eg electrostatic precipitators). In the temperature range> 200 ° C it comes through oxychlorination of these soot particles to form polychlorinated dibenzo-p-dioxins and -furans (PCDD / F) by the so-called de-novo synthesis. The particulate carbon (soot particles) is the dominant carbon source. The PCDD / F education takes place even with short-term disturbances over a very long period of time. The maximum PCDD / F formation depends on the amount of soot deposition rate. Even if the firing process is controlled again, the PCDD / F formation still takes place as long as carbon particles are present in the boiler deposits (memory effect).

Such disturbances can be detected by the aforementioned real-time measurements of the local soot concentration and deliberately reduced / avoided by a timely regulated air supply and intensive mixing in the region of the exhaust gas burnout zone.

Furthermore, the invention generally allows the particulate emissions (soot particles) of combustions, in particular of inhomogeneous fuels, to be detected and effectively reduced on the basis of control variables derived therefrom.

Literature:

1. [1] DE 103 47 340 A1
2. [2] http://www.weblearn.hs-bremen.de/risse/AWI/TEXTUR/merkmale.htm, as of 13.09.2006

LIST OF REFERENCE NUMBERS

1

Low-carbon combustion area

2

grate area

3

Sooty area

4

Transition line

5

Transition segment

6

Stainless transition segment

7

Rußübergangssegment

8th

stainless segment

9

Rußsegment

Claims (3)

1. Method for characterising the exhaust gas burn-off quality of a combustion process in combustion systems having a gas burn-off zone,
   characterised in that
   in a flow cross-section of the gas burn-off zone, low-soot combustion regions, regions without combustion, and regions of soot in the visible wavelength range are optically sensed, wherein the regions without combustion and the regions of soot are distinguished by different dynamics, and, by evaluation of a plurality of successive individual recordings can be differentiated in their junction regions with the low-soot combustion regions.
2. Method according to claim 1, comprising the following method steps:

   a) provision of a camera carrying out assessment of the flow cross-section in the gas burn-off zone against the flow, for a visible wavelength range,

   b) subdivision of the flow cross-section into segments with in each case a number of image points or pixels,

   c) recording of the combustion in the flow cross-section by the camera, with at least two individual image recordings following one another in temporal succession,

   d) localisation of the low-soot combustion regions in the flow cross-section, in that an intensity value arises above an adjustable intensity threshold value, which is automatically or manually adjusted relative to the maximum intensity in the flow cross-section,

   e) localisation of the junction areas as the image points or pixels which were allocated to the low-soot combustion region, but which comprise at least one adjacent image point or pixel which does not belong to the low-soot combustion region,

   f) allocation of the junction areas to junction segments,

   g) determination of a contrast value for each junction segment,

   h) allocation of the junction segments with contrast values above a contrast threshold value to a junction between a low soot combustion region and a region without combustion, and below the contrast threshold value to a junction between a low soot combustion region and a region of soot,

   i) iterative allocation of all segments in which the intensity of more than half the image points or pixels lies below the intensity threshold value, to the region of soot or to the region without combustion, by evaluation of neighbourhood relationships to junction segments and already allocated segments,

   j) determination and localization of the regions of soot by the proportions and allocation of the segments allocated to the region of soot.
3. Method according to claim 2, characterised in that determination and localization of the regions of soot are converted into control signals for a location-differentiated introduction, adjusted to the local combustion state, of a gas containing oxygen into the exhaust gas burn-off zone.

Images