GB2059574A - Absorption cell gas monitor - Google Patents

Abstract

An absorption cell gas monitor for measuring and monitoring the concentration of a specific gas, such as ozone, in an ambient atmosphere comprises first and second absorption cells (30) and (34) through each of which is passed alternately an ambient sample from an inlet port (12) and containing the specific gas to be monitored and an ambient sample from which the specific gas has been removed by a chamber (14) connected to the inlet port (12), each cell receiving the sample from the inlet port (12) while the other cell receives the sample from the chamber (14). Radiation from a source (60) having a wavelength absorbed by the specific gas is directed through each cell (30), (34) and the intensity of the radiation transmitted by each cell is detected by a detector (66), (68) respectively, the intensities being compared and computed in a stepped sequence in a microcomputer (72) to provide output data on the concentration of the specific gas in the ambient sample. <IMAGE>

Description

SPECIFICATION Absorption cell gas monitor This invention relates to absorption cell gas monitors for monitoring the concentration of a specific gas or vapour in an ambient atmosphere, and is particularly concerned with the rapid cancellation in real time of first order errors caused by absorption of radiation by gases and/or vapours other than that being monitored and by variations in output intensity of the source of radiation, as well as second order errors caused by optical asymmetry and reflectivity changes.

The presence or concentration of a gas or vapour in a sample is frequently determined by its characteristic absorption or attenuation of radiation of a particular wavelength. For any given gas or vapour to be measured or monitored, the sample is irradiated by energy of a wavelength which is significantly absorbed by the gas or vapour being monitored, but not by other gases or vapours which are expected to be present in the sample, and the transmission loss suffered by the radiation is determined.

The range of wavelengths which may be used in absorption cell gas monitors is rather wide. For example, gases such as ozone and sulphur dioxide, and vapours such as acetone and benzene significantly absorb radiation of wavelengths in the untraviolet region, whereas wavelengths in the infrared region are readily absorbed by gases such as NO2, CO2 and H2S, and by water vapour.

The present invention is primarily concerned with ozone monitoring, but is also applicable to absorption cell gas monitors for other gases and vapours, provided appropriate selection of the radiation source and other components of the monitor is made.

In a basic arrangement for monitoring the presence or concentration of ozone or any other specific gas in a sample, a comparison is made between the transmission of radiation of a suitable wavelength through an absorption cell containing a sample from which the specific gas has been removed and the transmission of similar radiation through the cell when it contains a sample from which none of the specific gas has been removed.The theoretical basis for such arrangements is found in the Beer Lambert Law: I/lo = e(-kLC) where I = the intensity of radiation transmitted by a sample containing the specific gas; I O = the intensity of radiation transmitted by a sample having the specific gas removed; k = the absorption coefficient of the specific gas at the wavelength of the radiation; c = the concentration of the specific gas in the sample; L = the length of the absorption cell; and, e = the natural logarithmic base.

Generally, in known absorption cell gas monitors, lo is measured first, using a suitable photoelectric device, and then I is measured to determine the difference in transmission through the absorption cell, and the concentration of the specific gas in the sample is simply calculated using the above Beer-Lambert equation.

Because it is difficult to maintain a constant output from the source of radiation, which is usually a light source, it has been common to place a reference photoelectric detector in a position to view directly the light source, and the output of the reference detector is used to compensate for fluctuations in the output of the source.

Measurements made in the manner described have produced some useful results, but the accuracy of those results suffers for several reasons. For example, the sample may be passed through a suitable chamber containing a scrubber or other means for removing the specific gas, but other gaseous and vaporous components which absorb radiation at the particular wavelength may remain. If the concentration of such other components varies with time, a measurement of 1o made prior to that of I may no longer be valid when I is measured.

Also, in procedures in which a single absorption cell is used, in addition to the possible errors resulting from the time delay between the measurement of 1o and the measurement of I, data-gathering time is lost because it is necessary to purge or flush the cell with the sample from which the specific gas has been removed to establish a "zero" base before the measurements are made. Typically, one might spend as much as a half-minute in the process of purging, zeroing and taking a single measurement, and in a given half-minute less than 10% of the available time may be spent in actual measurement.

In another application which we are filing today but which claims priority from United States Application, Serial No. 051,651, we describe how a considerable improvement in error cancellation and updating of data acquisition can be achieved by utilizing a reference absorption cell between the source of radiation and a reference detector, passing a sample from which the specific gas has been removed continuously through the reference cell, and using measurements taken from both the reference cell and the test cell. In a preferred example two test cells are operated in conjunction with the reference cell, the test cells being operated on a 50% duty cycle 1 80 degrees out-of-phase with each other.However, each absorption cell is the same length as that used in the previously known absorption cell gas monitors for the same signal-to-noise ratio, and volume and flow rate are twice that of previous systems for comparable flush times.

According to the present invention, an absorption cell gas monitor for monitoring the concentration of a specific gas or vapour in an ambient atmosphere comprises means for continuously drawing a sample of the ambient atmosphere into an inlet port, a chamber connected to the inlet port and containing material which removes the specific gas or vapour from the ambient sample passing through the chamber, a source of radiation of a predetermined wavelength which is absorbed by the specific gas or vapour, first and second absorption cells each of which is disposed so that radiation from the source passes through it, first means for connecting the outlet from the chamber to the first and second cells alternately to cause periodic flows of the ambient sample from which the specific gas or vapour has been removed through the first and second cells, second means for periodically connecting the inlet port to the second and then the first cell to cause an intermittent flow of the ambient sample through each cell which is out-of-phase with the periodic flow through the cell of the ambient sample from which the specific gas or vapour has been removed, and means for detecting and comparing the absorption of radiation from the source by the contents of the first and second cells in a predetermined sequence.

In the monitor in accordance with the present invention, the basio absorption cell theory described earlier is followed. However, no separate reference absorption cell is used as in the monitors described in our United States application No. 51,651. Instead, two test cells are alternately supplied with sample containing the specific gas and sample from which the specific gas has been removed in a completely out-of-phase relationship, and cancellation of the effect of gases or vapours other than the specific gas that absorb radiation at the chosen wavelength in real time is achieved by a comparison of absorption of radiation by the two cells on a stepped basis. Errors resulting froth variations in radiation source intensity are minimized by integrating simultaneously the outputs of both cells in a first and then in a second mode.Also, a minimum flush time between the alternate measurements is maintained and output information on the ambient concentration of the specific gas can be updated every half-cycle of operation.

An assumption is made that source output intensity changes are the same for both absorption cells and with a short time lapse between measurements and a symmetrical optical system accuracy of measuremerits is assured. Furthermore, with the monitor in accordance with the invention, the cell lengths and gas flow rates need only be half those required in the monitors described in our United States Application No. 51,651 for a given signal te noise ratio.

An example of an absorption cell gas monitor in accordance with the invention for monitoring the concentration of ozone in an ambient atmosphere will now be described with reference to the accompanying drawing which is a schematic diagram of the monitor.

In Figure 1, a port 12 for the sample intake from the ambient atmosphere is shown communicating directly with a chamber 14 which contains an ozone-removing catalyst. The ozone removing catalyst may be in the form of a fixed bed filter of metallic oxides, although other ozone removing systems may be used. The inlet port 12 2 is also directly connected to a normally closed inlet port 1 6 of a solenoid valve 1 8 and to a normally closed inlet port 20 of a solenoid valve 22. The outlet of the chamber 14 communicates with a normally open second inlet 24 of the valve 1 8 and with a normally open second inlet 26 of the valve 22. With such an arrangement the system is in a mode to flush both cells 30 and 34 with ozone-free sample gas.

An outlet line 28 runs from the outlet of the solenoid valve 1 8 to an absorption cell 30. A similar line 32 connects the outlet of the solenoid valve 22 to an absorption cell 34. An outlet line 36 from the absorption cell 30 runs to a flow meter 38 and through a needle valve 40 to a pump 42 to an exhaust port 44. Similarly, an outlet line 46 from the absorption cell 34 runs to a flowmeter 48 through a needle valve 50 to the pump 42 and the exhaust port 44. The pump 42 continuously exhausts the contents cf the system through the exhaust port 44. The individual needle valve and flowmeter combinations permit close control of the flow through the system. The various solid lines and the elements described constitute a simplified showing of the mechanism involved in the flow of gas through the apparatus of the monitor.

In addition to the solid lines indicating the gas flow through the apparatus, dotted lines show the path of radiation, in the form of light in the case of ozone monitoring. The light is generated preferably by a low-pressure mercury vapor lamp 60, 95% of the total output of which is concentrated in a line at 254 nm. The light is trained upon a first mirror 62 and a second mirror 64. From the mirror 62, light is reflected through the length of the absorption cell 30 to a detector 66. The detector 66 may be any of several known devices such as a solar-blind vaccum photodiode detector sensitive to 254 nm radiation.

The output current of the detector 66 is determined by the intensity of radiation of the preselected wavelength impinging upon the detector. In this fashion, an efficient monochromator for the detection of ozone is formed. At the same time, light from the source 60 is reflected by the mirror 64 through the length of the absorption cell 34 to a detector 68 which may be similar in all respects to the detector 66, a second monochromator being thus formed.

In addition to the solid lines indicating gas flow and dotted lines indicating light paths, dashed lines show the path of electrical signals. A signal in the form of current is derived from the detector 66 and is coupled to a reference digital electrometer 70 from which it is passed to a microcomputer unit 72 which may generally correspond to the comparable unit disclosed in the cited parent application Serial No.051,651. That unit, as noted, includes a standard 8-bit microprocessor and other elements described below. At the same time as the signal is derived from the detector 66, a signal is also generated in the detector 68, coupled to a digital electrometer 74 and also passed to the unit 72.

Besides providing data for display or other use at an output terminal 76, the microprocessor of element 72 is programmed to provide a suitable output to control the operation of the solenoid valves 1 8 and 22 as described below.

The signals derived from the detectors 66 and 68 may be processed by first being converted to voltages and then to frequencies by voltage-to-frequency converters. The output of the converters may then by counted by counters for a specific length of time determined by the microprocessor 72 and integrated in the manner described below.

In operation, a cycle of the ozone monitor begins as the solenoid valve 18 is switched to route the sample from the ozone-removing chamber 1 4 into the cell 30 (absorption cell A). At the same time, the solenoid valve 22 is switched so as to route the sample directly from the port 12 into the cell 34 (absorption cell B). After an appropriate flush time, typically 4-5 seconds, 1o for the cell 30 is determined, and I for the cell 34 is detemined.

The intensities of the currents from the detectors 66 and 68 may be integrated over a period of one second, and these values are stored in the memory of the microcomputer 72. These values for the first integration period may conveniently be denoted lo(A,t=1) and l(B,t=1) for the first half of a cycle.

At the end of the first integration period, a signal from the microcomputer 72 triggers the solenoid valves 1 8 and 22 so that sample passes directly from input 12 through the solenoid 1 8 to the cell 30 and simultaneously the ozone-free output of the chamber 14 passes through the solenoid 22 to the cell 34. After a flush of about 5 seconds, the intensities of the currents from the detectors are integrated once more over a period of one second. During this second half of the cycle, the values measured for this second integration period may be denoted l(A,t=2) and lo(B,t=2).

The Beer-Lambert equations for times t=1 and t=2 are:

where L(A), L(B) are the lengths of absorption cells A and B, c(t=1) and c(t=2) are the ozone concentrations at times t=1 and t=2, l0(B,t=1) is the photocurrent that would have been measured by detector B during the time t=1 if the sample flowing during time 1 had its ozone removed, and l0(A,t=2) is similarly defined photocurrent for detector A at time t=2.

As is clear from their definition, the values lo(B,t=1) and íO(A,t=2) cannot be measured directly. If the stability of the lamp system in the detectors from time t=1 to time t=2 were sufficient, we could simply set lo(B,t=1), lo(A,t=2) equal to the measured values lo(B,t=2) lo(A,t=1). In practice, sufficient stability cannot be obtained, primarily because of variation of the lamp intensity with time IstaSififies of one part in 105 are needed for detection of ozone concentration of one part per billion). Thus, the simple assumptions lo(A,t=2) = lo(A,t=1), lo(B,t=1) = lo(B,t=2) cannot be used. Let d(A) and d(B) be the factors expressing the change in lamp intensity between time t= and time t=2, as measured by detectors A and B, respectively.We will have: I =(A,t=2) = d(A) lo(A,t=1) and lo(B,t=2) = d(B)lo(B,t=1), (3) 1 that is, lo(B,t=1) = lo(B,t=2) (4 d(B) Substituting (4), (3) in (1) and (2) we obtain:

Multiplying these two equations gives

If one assumes: A) L(A)=L(B)=L and: B) d(A) = d(B) and defines: C) c = [c(t=1) + c(t=2)]/2 one obtains:

where cis the only unknown.

Assumption A is valid by virtue of the construction of the system. Assumption B - that is, that the total intensity changes are identical for channels A and B - is valid to the first order. However, certain second order effects, such as changes in reflectivity and in the spatial distribution of radiation, may not be identical for each channel.

As is apparent from the exponent of equation (5), the effective cell length is 2L, so that the required length for each individual cell 30 and 34 is half that of the invention disclosed in the above cited application, Serial No. 051,651 for the same signal-to-noise ratio. With a typical cell length of about 40 cm. for cells 30 and 34, these cells need not be folded, and hence internal mirrors are not required. Since the d's of assumption B are a function of lamp intensity changes and changes in the internal reflective properties of the cells, a symmetrical optical system and the absence of reflective components due to the elimination of internal mirrors tends to minimize possible second order errors.

The short cell length also permits faster flushing of the cells for a given rate of gas flow than for the cells described in application Serial No. 051,651. This reduces the time interval between the determination of intensities at time period 1 and time period 2 and increases the validity of assumption B. It should also be noted that less lag time is encountered in updating output information, since c is determined in stepped fashion from intensities measured first at times one and two, then at times two and three, etc.

Thus, new information is available every half cycle.

Effective cancellation of errors caused by substances other than ozone that absorb light at 254 nm is also achieved. This can be illustrated as follows: If there is a concentration c1 with an absorption coefficient of k1 of substances other than ozone at time t=1, then all intensities measured will be reduced by exp[--k, cj(t=1) Lj, while at t=2 all intensities are reduced by exp[--k, c,(t=2) L].

If the reduced intensities are substituted into equation (5), the following expression is obtained:

Since this expression reduces to the equation (5) above, it is seen that the effects of absorption by substances other than ozone cancel in the system of the present invention here as in the device of the prior patent application.

As in the above-identified patent application Serial No. 051,651, pressure and temperature transducers are employed. A pressure transducer 80 which may be a commercially available hybrid l.C.

laser-trimmed strain gauge having internal temperature compensation and stability of the order of 1% of full scale output is located immediately adjacent the exit outlet of the absorption cell 30 which connects to the outlet line 36. Its voltage output is, of course, fed to the microcomputer of element 72 as in the apparatus of the parent application.

Similarly, a temperature transducer 82 produces a signal which is fed into the microcomputer of element 72. The temperature transducer is located at approximately the midpoint of the absorption cells and it may be a commercially available device laser-trimmed to produce 298.2 micro-amps output at 298.20 K (250C). Computations are carried out in a manner similar to that of the parent application, Serial No. 051 ,65 1, insofar as the interfacing of the pressure transducer and temperature transducer are concerned.

The operation of the microcomputer 72 is in fact entirely similar to that of the comparable element in the parent application, Serial No.051,651 except for the signals which are provided to solenoids 18 and 22. Obviously, the switching of these solenoids to alternate the flow of ambient and ambientminus-ozone in the manner described above is easily accomplished but follows a somewhat different sequence from the switching of the solenoids or solenoid valves of the parent application.

Claims (6)

1. An absorption cell gas monitor for monitoring the concentration of a specific gas or vapour in an ambient atmosphere, comprising means for continuously drawing a sample of the ambient atmosphere into an inlet port. a chamber connected to the inlet port and containing material which removes the specific gas or vapour from the ambient sample passing through the chamber, a source of radiation of a predetermined wavelength which is absorbed by the specific gas or vapour, first and second absorption cells eachof which is disposed so that radiation from the source passes through it, first means for connecting the outlet from the chamber to the first and second cells alternately to cause periodic flows of the ambient sample from which the specific gas or vapour has been removed through the first and second cells, second means for periodically connecting the inlet port to the second and then the first cell to cause an intermittent flow of the ambient sample through each cell which is out-of-phase with the periodic flow through the cell of the ambient sample from which the specific gas or vapour has been removed, and means for detecting and comparing the absorption of radiation from the source by the contents of the first and second cells in a predetermined sequence.

2. A monitor according to claim 1, in which the first means for connecting the outlet of the chamber to the first and second cells alternately and the second means for periodically connecting the inlet port to the second and then the first cell comprise a pair of three-way solenoid valves having two inlets and one outlet, and computer means for providing signals which actuate the solenoid valves in a sequence related to the predetermined sequence.

3. A monitor according to claim 1 or claim 2, in which the specific gas to be monitored is ozone, the material in the chamber comprising metallic oxides for removing ozone from the ambient sample which passes through the chamber, and the means for continuously drawing ambient sample into the inlet port is a continuously operating pump connected to the absorption cells downstream thereof.

4. An absorption cell gas monitor for monitoring the concentration of ozone in an ambient atmosphere, comprising means for continuously drawing a sample of the ambient atmosphere into an inlet port, a chamber having an inlet connected to the inlet port and containing material for removing ozone from the ambient sample which passes through the chamber, a source of radiation having a wavelength of substantially 254 nanometers, first and second absorption cells each disposed so that radiation from the source passes through it, first and second three-way solenoid valves each having an inlet connected to the inlet port and another inlet connected to the outlet of the ozone removing chamber, the outlets of the first and second valves being connected to the first and second cells respectively, a microcomputer electrically connected to the first and second valves for controlling their operation in accordance with a predetermined program periodically to cause flow of the ambient sample from which ozone has been removed to the first cell and flow of the ambient sample from the inlet port to the second cell followed by flow of the ambient sample from which ozone has been removed to the second cell and flow of the ambient sample from the inlet port to the first cell, and detection means for detecting the absorption of radiation from the source by the contents of the first and second cells, the detection means being connected to and providing inputs to the microcomputer for storage and computation thereof in a predetermined sequence to provide output data on the concentration of ozone in the ambient sample.

5. A monitor according to claim 4, in which the first and second solenoid valves are controlied by the microcomputer so that the first cell is alternately connected to the chamber and to the inlet port, and the second cell is alternately connected to the inlet port and to the chamber 1800 out-of-phase with the first cell.

6. A monitor according to claim 1, substantially as described with reference to the accompanying drawing.