GB2261944A

GB2261944A - Flame monitoring apparatus and method

Info

Publication number: GB2261944A
Application number: GB9123971A
Authority: GB
Inventors: Brian Glassby
Original assignee: National Power PLC
Current assignee: National Power PLC
Priority date: 1991-11-12
Filing date: 1991-11-12
Publication date: 1993-06-02
Also published as: GB9123971D0

Abstract

An apparatus and method are provided for discriminating between flames from different fuel sources. Such an apparatus and method may be used in multiburner installations to discriminate between flames from oil and flames from pulverised fuel. In the apparatus and method, a characteristic of the variation of the amplitude of flicker of radiation from the system with flicker frequency is detected. Although there is no particular frequency peculiar to one source of fuel, by which flames from that source of fuel can be identified, it is possible to make use of the different spectra of flicker frequencies produced in different burner operating conditions to discriminate between flames from different sources of fuel. <IMAGE>

Description

FLAME MONITORING APPARATUS AND METHOD This invention relates to monitoring flames from different fuel sources in a way that enables discrimination between flames from one fuel source and flames from another. In particular, the invention relates to an apparatus and method for discriminating between flames from oil and flames from pulverised fuel (PF), in for example power station multiburner installations where such flames may be present at the same time.

A typically multiburner installation for a coal-fired boiler may include rows of six PF burners at five levels on the boiler, front and rear. PF burners cannot be ignited from cold and so an oil burner is provided substantially coaxially with each PF burner.

Viewing ports are provided for the combination each PF and oil burners so that the operation of the burners can be monitored.

A considerable amount of effort over many years has been devoted to the pursuit of reliable flame monitoring, by power station operators, manufacturers of burner equipment, and specialist instrument organisations.

Whilst detection of the electromagnetic emissions from flames over the range ultraviolet (100 to 300nm) through the visible, (400 to 600nm) to the infra-red (6C0 to lO,OOOnm) is readily achieved by a variety of transducers, the ability to detect specific flames from a mass of similar flames has proved elusive. The cross correlation technique has been a successful and reliable means of monitoring main PF flame presence, and ensuring safe furnace operation. However, the requirements for oil light up burners are more difficult to meet since the oil flames are smaller with restricted viewing access, and must be detected against the bright PF flame background under a wide range of conditions from first start on a black boiler to PF support at maximum boiler load.

Accurate and reliable monitoring of oil flames is essential in mixed fuel burners in order to achieve the level of safety required. CEGB standard GD and CD166 and BS799 require that before attempting to introduce an oil burner, a check should be made to ensure that there is no oil flame present, and during operation that loss of an oil flame should be detected to allow the monitored burner to be shut down immediately. So far, it has not been possible to meet these requirements consistently with most of the flame monitoring equipment used. The main reason for this is the inability of the equipment to discriminate between background PF flames and oil flames.

US4370557 (Axmark et al - assigned to Honeywell Inc.) discloses a flame monitor comprising two radiation sensors. One sensor is a silicon (Si) detector which is sensitive to visible light. The other sensor is a lead sulphide (PbS) detector which is sensitive to infra-red (IR) light. The monitor exploits the fact that flames of multiburner arrays all characteristically have both a DC and an AC (or flicker) component in radiation intensity.

The AC components of the various flames in the boiler generally cancel out so that the only significant AC component to be observed comes for the nearest flame, ie.

the flame under observation. The AC signals from the two sensors are summed and the DC signal from the Si sensor is used to control the gain applied to the AC signal. If the burner being observed is not operating but a number of other burners are operating, there will be a background fireball and the AC component seen by the fuel detector of the burner being observed will be substantially non-existent. When the oil fed pilot torch is inserted into the burner being observed, the oil flame provides a strong AC component signal in the visible range which is seen by the Si sensor so the indicator shows a safe condition. When the main PF flame is ignited and burning, the magnitude of both the AC and DC signal will decrease in the visible range but the AC signal will increase in the IR range to again indicate safe conditions.In this monitor, the flicker characteristics of the flames in both the visible and the IR spectrum is claimed to be significant.

An object of the present invention is to provide a flame monitor capable of distinguishing between oil flames and flames from other sources such as pulverised fuel.

One aspect of the present invention provides apparatus for distinguishing one flame produced by one source of fuel in a system including another source of fuel for producing another flame, the apparatus comprising means for detecting a characteristic of the variation of the amplitude of flicker of radiation from the system with flicker frequency.

The flicker frequency of a flame is the very low frequency at which radiation from the flame is modulated, and ranges from about 2Hz to 500Hz.

The inventor has appreciated that, although there is no particular frequency peculiar to one source of fuel, by which flames from that source of fuel can be identified, it is possible to make use of the different spectra of flicker frequencies produced in different burner operating conditions. In particular, it was found that, for a PF and oil burner, the presence of an oil flame in the burner under observation decreased the amplitude of flicker at low flicker frequencies and increased the amplitude of flicker at high frequencies.

This approach may be contrasted with the approach of the prior art monitor of US4370557 in which the amplitudes of flicker produced by radiation of different wavelengths from the flame (the visible spectrum and the infra-red spectrum) is considered.

A second aspect of the present invention provides a method of distinguishing one flame produced by one source of fuel in a system including another source of fuel for producing another flame, the method comprising the steps of detecting a characteristic of the variation of the amplitude of flicker of radiation from the system with flicker frequency.

The method may include the initial steps of determining a cross-over frequency for the system at which the amplitude of flicker of radiation due to said one flame is the same as the amplitude of flicker of radiation due to said another flame; and choosing said first flicker frequency of frequency band to be less than said cross-over frequency and choosing said second flicker frequency or frequency band to be greater than said cross-over frequency.

These optional steps in the method allow for the variation in amplitude of flicker of radiation from the system with flicker frequency to be a characteristic of the system.

Examples of embodiments of the present invention will now be described with reference to the drawings in which: Figures 1, 2 and 3 are graphs showing the difference in flicker frequency spectra between flicker due to background PF radiation and flicker due to an oil flame for different mixed fuel burners; Figure 4 is a schematic graph showing a variation in flicker frequency spectra between flicker due to background PF radiation and flicker due to an oil flame which may arise for another mixed fuel burner; Figure 5 is a block diagram of one embodiment of the present invention;; Figure 6 shows a circuit for producting two DC signals corresponding to the detected amplitude of modulated radiation at two discrete frequency bands in accordance with a preferred embodiment, and Figure 7 shows a circuit for computing the ratio of the two DC signals, and for providing both display and control signals.

Figure 1 shows, for one mixed fuel burner, the variation of flicker amplitude with flicker frequency for an oil flame against a background of PF radiation. When the oil flame is present, it generally fills the monitor view obscuring the low PF flames. In this example, the flicker amplitude for the background is greater than that for oil for frequencies below about 60Hz and is less than that for oil for frequencies above about 75Hz, ie. there is a cross-over between 60 and 75Hz. By computing the ratio of amplitude of flicker at frequencies above and below the cross-over it is possible to determine if the oil pilot flame is lit even in the presence of background PF flames. This ratiometric technique has the advantage of being amplitude independent over the normal working range, thus giving immunity from the wide signal level variations that exist.

In addition, a significant difference exists in the characteristics of low and high frequency energy in respect of crest factor, the peak to RMS ratio. This occurs because the background PF radiation received from a turbulent flame skirt is a mix of several uncorrelated background flames of slightly different flicker frequencies even. In contrast, the oil flame under observation generally fills the monitor view obscuring the background flames, and being a single flame provides a signal more closely approximating to a single frequency.

Higher frequencies tend to predominate in the oil flame because air and fuel mass flows are generally smaller than those for the PF burners.

Figure 2 shows the variation of flicker amplitude with flicker frequency for an oil flame against a background of PF radiation in another mixed fuel burner in a multiburner installation. In this example, the cross-over of the curves for the two situations (PF background only and oil flame present) occurs at about 1OHz. Thus, it is possible to determine if the oil pilot flame is lit even in the presence of background flames by computing the ratio of the amplitude of flicker at frequencies above and below the cross-over.

Figure 3 shows the variation of flicker amplitude with flicker frequency in yet another mixed fuel burner.

The flicker amplitude detected for the PF flame is greater than that for oil at frequencies below about 20Hz and is less than that for oil for frequencies above about 20Hz, ie. there is a cross-over at about 20Hz. By computing the ratio of amplitude of flicker at frequencies above and below the cross-over, it is possible to determine if the oil pilot flame is lit even in the presence of PF flames.

Another possible variation of flicker amplitude with flicker frequency for an oil flame against a background of PF radiation which may occur in certain mixed fuel burners is shown in Figure 4. Cross-over of the curve for background flicker and the curve for the oil flame occurs at several frequencies. However, the two curves can be differentiated by detecting the amplitude of flicker at low and high frequencies in different ways. At low frequencies, the amplitude detected by the monitor can be the peak amplitude. If the peak amplitude is detected at low frequencies, then it is sufficient to detect the average amplitude of flicker at high frequencies above 200Hz. Discrimination between the oil flame and the background flames may be less evident for mixed fuel burners with this characteristic than for the mixed fuel burners of Figures 1, 2 and 3.However, such discrimination may be sufficient for the monitor of the present invention to be a useful monitor.

From the above, it will be appreciated that the phenomenon, on which the present invention relies, can vary quantitatively from burner to burner. The variation is at least in part due to the construction of the mixed fuel burner and the relative positions of the oil burner, the PF burner and the monitor.

One embodiment of a monitor for a mixed fuel burner uses a single radiation sensor aimed at the mixed fuel burner. A signal representative of flicker in the radiation from the mixed fuel burner is produced by the sensor. This signal is processed by a circuit to determine the amplitudes of flicker of radiation at two flicker frequencies or frequency bands which are compared to provide an indication of the variation of amplitude of flicker of radiation from the burner with flicker frequency.

An example of circuitry for use in the monitor of the present invention is shown in block form in Figure 5 and in greater detail in Figures 6 and 7. Connections between the circuitry of Figure 6 and the circuitry of Figure 7 are indicated by the arrows X, Y and Z. Visible and near infra-red radiation detected by a transducer 1 such as, for example, a silicon photodiode is converted into an electrical analogue signal 2 corresponding to the flame instantaneous intensity in the visible and near infra-red spectrum. This signal appears at an input 3 which is referenced to input 5 of an amplifier 7. The amplifier 7 produces a preamplified signal 8, with the required gain set by a trimpot 9. The input CR circuit 10 ensures that frequencies from DC to approximately 2Hz are attenuated. The analog signal them passes to a high pass filter 11 which attenuates signal frequencies above 16Hz.

The attenuated signal is then passed to a full wave rectifier 15 whose gain is adjustable by a level control 16. The output of the full wave rectifier 15 is passed to a store 17 at which the peak value of the low frequency waveform is stored with a delay time constant to provide a slow delay on reduction of signal level. A buffer 19 is connected to the store 17 and forms V1, the DC zero level being adjustable. In this way, the circuit produces a signal V1 representative of the peak or maximum amplitude of flicker of radiation from the system at a first band of flicker frequencies in the range of from 2Hz to 16Hz.

Referring to the lower circuit of Figure 6, the signal 8 from the amplifier 7 is also passed to a double high pass filter 21 formed by IC4A and IC4B, which reaches unity gain at a frequency of 210Hz, rapidly attenuating signals below this value. This signal is then rectified by full wave rectifier 23 formed by IC5A and IC5B, with the gain adjustable by means of the level control 23a.

The signal is then averaged by filter 25 formed by R28, C15 and R29 and is buffered by buffer 27 to produce a second DC signal V2, a DC zero level being adjustable by VR5. In this way, the circuit produces a signal V2 representative of the average amplitude of flicker of radiation from the system at a second band of flicker frequencies above 210Hz.

The two DC signals V1, V2 are proportional to the frequency content (the peak amplitude or the average amplitude of flicker) in the selected frequency bands.

These signals V1, V2 are passed, as indicated by arrows Y and Z, to a divider 29 formed by IC7 as shown in Figure 8. At divider 29, the quantity

V = V2 x 10 volts OUt V1 is computed. This ratiometric approach means that the value of V out is essentially independent of the absolute values of the DC signals over a range of 50mV to 10V and ensures that circuit operation is unaffected by wide variations in the input signal level as furnace operating conditions vary.

An offset voltage is provided by R34 and VR6 to enable the divider 29 to be cut off below a selected input level. This compensates for the case where V1 is zero (on black boiler start up conditions), and prevents spurious output in the event of loss of the flame monitor head.

The output of the divider 29 is displayed on a small panel meter 33 (shown in Figure 6) and is also passed to comparator 31 formed by IC6, which holds in relay RL/2 via transistor Q1 when the output level exceeds half scale, and energises LED34 to indicate an oil flame in "ON" status. A volt-free changer contact and a contact closed for "oil flame on" are available to a burner management system.

As described, the monitor makes use of the frequency response of the amplitude of flicker of radiation from the mixed fuel burner to determine whether or not an oil flame is present. In the example described, the characteristic detected by the monitor is the relative change of the amplitude of flicker of radiation from the system at different frequencies depending on the presence or absence of an oil flame in the burner under observation.

During initial start up, the oil burner must be lit first. However, some ratios of oil and air are explosive so it is necessary to. determine whether or not the oil flame is present before any attempt is made to vary the amount of oil in the system. The monitor of the present invention was found to allow the absence of an oil flame to be checked, even against a background of PF flames from other burners, before any attempt is made to ignite the oil burner.

Ignition of the PF burner requires the presence of an oil flame. The monitor of the present invention was found to allow the presence or absence of the oil flame to be detected against a background of PF flames from other burners.

Once the PF burner has been lit, the oil burner can be maintained lit or extinguished depending on the required management of the mixed fuel burner in the multiburner installation. Sometimes the oil burner is not required and so can be extinguished. However, if an oil flame is required, at any stage, for the proper operation of the mixed fuel burner, loss of the oil flame may create a hazardous situation and so must be detected to allow the monitored burner to be shut down immediately. The monitor of the present invention was found to allow the presence or absence of an oil flame to be detected against the background of PF flames from the mixed fuel burner.

When the monitor is installed on site, the gain of the various parts of the circuit of Figures 6 and 7 needs to be correctly set. Initially, the PF burner is set up at significant boiler load with no oil flame present. The radiation sensor is aimed towards a bright part of the PF flames and the meter 33 is connected to represent the signal V1 via a set PUSH BUTTON 35 and the control 16 is adjusted to show an 80% reading of full scale. The oil burner is then established and the meter is connected to Vout (by not operating the push button 35) and the reading of the meter 33 set for 90% of full scale using the control 23a.

The monitor of the present invention has been tested with a number of mixed fuel burner installations in which the flames from the burner under observation can be viewed along an axis of the oil burner, an unobstructed view of the furnace being available. In front wall fired units, the view of the oil flame is down an air tube concentric with the oil burner and through the back of swirler blades used to control movement of air in the oil burner. This view allows the full PF background levels to be observed until the oil burner is introduced, when the oil flame envelope is seen instead.Tests at a number of mixed fuel multiburner installations showed that the monitor of the present invention provided a significant improvement in the discrimination of the oil flame, in some cases producing good results where discrimination by prior art monitors had been non-existent or virtually non-existent. For some mixed fuel burners, the gain of the circuit may need to be increased or the angle of view of the radiation sensor modified to be aligned on the brightest part of the oil flame to provide sufficient discrimination. Those skilled in the art will appreciate that, as the amplitude of flicker due to the background flames may also vary transiently, it may be necessary, in some cases, to modify the response time of the monitor so that transient changes in the background flames alone do not provide a spurious indication of the presence or absence of an oil flame. Tests have also showed that equivalent performance can be achieved from an angled viewing position aimed at the oil flame route.

Embodiments of the present invention may also provide assistance in discriminating the presence of an oil flame in the burner being monitored against a background of oil flames from adjacent burners. The background oil flames would produce radiation which is a mix of several uncorrelated background flames of slightly different flicker frequencies. In contrast, an oil flame in the burner being monitored would generally fill the monitor view, obscuring the background flames, and being a single flame producing a signal more closely approximating a single frequency.

Modifications to the embodiments described hereinbefore will be apparent to those skilled in the art.

Claims

1. Apparatus for distinguishing one flame produced by one source of fuel in a system including another source of fuel for producing another flame, the apparatus comprising means for detecting a characteristic of the variation of the amplitude of flicker of radiation from the system with flicker frequency.

2. Apparatus according to Claim 1 wherein said means for detecting comprises first detection means for detecting the amplitude of flicker of radiation from the system at a first flicker frequency or frequency band.

3. Apparatus according to Claim 2 wherein said first detection means comprises a radiation sensor for sensing radiation from the system and first processing means for processing the output of said sensor to produce a first signal representative of the amplitude of flicker of radiation from the system at said first flicker frequency or frequency band.

4. Apparatus according to Claim 3 wherein the radiation sensor comprises a silicon photodiode.

5. Apparatus according to Claims 3 or 4 wherein said first detection means includes means for detecting the maximum amplitude of flicker of radiation.

6. Apparatus according to any one of Claims 2 to 5 wherein said first frequency band comprises frequencies in the range of from 2Hz to 16Hz.

7. Apparatus according to any one of Claims 2 to 6 further comprising: second detection means for detecting the amplitude of flicker of radiation from the system at a second flicker frequency or frequency band; and means for comparing the detected amplitudes of flicker at said first and second flicker frequencies or frequency bands.

8. Apparatus according to Claim 7 wherein said first and second detection means comprise a common radiation sensor for sensing radiation from the system, first processing means and second processing means for processing the output of said sensor to produce a signal representative of the amplitude of flicker of radiation from the system at the first and second flicker frequencies or frequency bands.

9. Apparatus according to Claim 8 wherein said first and second processing means include respectively a first and a second filter means for determining the first and second flicker frequency bands.

10. Apparatus according to Claim 9 wherein said first flicker frequency band is less than said second flicker frequency band.

11. Apparatus according to any one of Claims 7 to 10 wherein said second detection means includes means for determining the average amplitude of flicker of radiation over said second flicker frequency band.

12. Apparatus according to any one of Claims 7 to 11 wherein the comparing means comprises means for determining the ratio of the detected amplitudes of flicker at said first and second flicker frequencies or frequency bands.

13. Apparatus according to any one of the preceding Claims 7 to 12 wherein said second frequency band has a lower limit of 210Hz.

14. A method of distinguishing one flame produced by one source of fuel in a system including another source of fuel for producing another flame, the method comprising the step of detecting a characteristic of the variation of the amplitude of flicker of radiation from the system with flicker frequency.

15. A method according to Claim 14 comprising the step of detecting the amplitude of flicker of radiation from the system at a first flicker frequency or frequency band.

16. A method according to Claim 15 comprising the step of detecting the maximum amplitude of flicker of radiation from the system in said first frequency band.

17. A method according to Claims 14 or 15 wherein said first frequency band comprises frequencies in the range of from 2Hz to 16Hz.

18. A method according to any one of Claims 14 to 17 including the further steps of: detecting the amplitude of flicker of radiation from the system at a second flicker frequency or frequency band; and comparing the detected amplitudes of flicker at said first and second flicker frequencies or frequency bands.

19. A method according to Claim 18 wherein the steps of detecting the amplitude of flicker of radiation from the system at a first and second flicker frequencies or frequency bands comprises the steps of: sensing radiation from the system to produce a common signal; and processing the common signal to determine the amplitudes of flicker of radiation from the system at the first and second flicker frequencies or frequency bands.

20. A method according to Claim 19 wherein the step of processing the common signal includes the step of filtering the common signal.

21. A method according to Claim 20 wherein said first flicker frequency band is less than said second flicker frequency band.

22. A method according to any one of Claims 18 to 21 wherein the step of detecting the amplitude of flicker of radiation from the system at a second flicker frequency band includes the step of determining the average amplitude of flicker of radiation at said second flicker frequency band.

23. A method according to any one of Claims 18 to 2.2 wherein the step of comparing the detected amplitudes comprises the step of determining the ratio of the detected amplitudes of flicker at said first and second flicker frequencies or frequency bands.

24. A method according to any one of Claims 18 to 23 wherein said second frequency band has a lower limit of 210Hz.

25. A method according to any one of Claims 18 to 24 including the initial steps of determining a cross-over frequency for the system at which the amplitude of flicker of radiation due to said one flame is the same as the amplitude of flicker of radiation due to said another flame; and choosing said first flicker frequency or frequency band to be less than said cross-over frequency and choosing said second flicker frequency or frequency band to be greater than said cross-over frequency.

26. A method according to any one of Claims 14 to 25 for distinguishing an oil flame produced by an oil burner in a system including at least one burner selected from the group of pulverised fuel burners and background oil burners.

27. An apparatus for distinguishing one flame produced by one source of fuel in a system including another source of fuel for producing another flame substantially as hereinbefore described with reference to any one of the accompanying drawings.

28. A method of distinguishing one flame produced by one sourve of fuel in a system indicating another source of fuel for producing another flame substantially as hereinbefore described.