Concentration measuring apparatus and method

Abstract

The invention is a concentration measuring apparatus for measuring a concentration (Y) of a component to be measured in a fluid flow that comprises a feeding unit and a detector unit coupled to the feeding unit. The apparatus according to the invention has a first detector unit (11 ) and a second detector unit (12), the feeding unit comprising a first fluid splitter unit (21 ) having an inlet configured to receive an examined volume flow of the fluid flow and a first and a second outlet configured to respectively output, in a first split ratio (AE), a first partial volume flow (Xi) and a second partial volume flow (X2), wherein the first outlet of the first fluid splitter unit (21 ) is in fluid communication with the first detector unit (11 ), and the second outlet of the first fluid splitter unit (21 ) is in fluid communication with the second detector unit (12), and a second fluid splitter unit (22) having an inlet configured to receive a volume flow of an additive fluid containing a known concentration (c) of the component and a first and a second outlet configured to respectively output, in a second split ratio (AA), a first partial volume flow (F1) and a second partial volume flow (F2) different from the first partial volume flow (F1). wherein the first outlet of the second fluid splitter unit (22) is in fluid communication with the first detector unit (11 ), and the second outlet of the second fluid splitter unit (22) is in fluid communication with the second detector unit (12). The invention is further a concentration measuring method implemented applying the apparatus for measuring a concentration (Y) of a component to be measured in a fluid flow.

Description

[0001] CONCENTRATION MEASURING APPARATUS AND METHOD

[0002] TECHNICAL FIELD

[0003] The invention relates to a concentration measuring apparatus for measuring a concentration of a component to be measured in a fluid flow, comprising a feeding unit and a detector unit coupled to the feeding unit. The invention also relates to a concentration measuring method implemented applying the apparatus.

[0004] The invention basically relates to the field of concentration-measurement instruments (i.e., analytical instruments) that operate on the so-called on-line measurement principle, i.e., the fluid flow, that is, the gas or liquid flow to be measured is being sampled continuously or quasi-continuously (or even intermittently), in the form of a partial flow, the concentration of the component or components to be measured being measured in a partial flow fed to the analytical apparatus.

[0005] BACKGROUND ART

[0006] One of the most important characteristics of measurements - for example, gas concentration measurements applying the photoacoustic principle or another spectroscopy-based measurement principle - is the signal-to-noise ratio (S / N) of the measurement, which is defined as the ratio of the so-called useful signal (S) generated by the component to be measured, and the noise (N) of the measurement. S / N usually has a complex relationship with the concentration of the measured component; however, in most cases a preferable S / N range envisaged for performing the measurements can be determined. It is preferable to perform the measurements such that the signal-to-noise ratio falls into this S / N range, because in this range one or more of the following statements are typically valid or approximately valid:

[0007] The measured signal (S) is proportional to the measured concentration.

[0008] The noise (N) of the measurement is independent of the measured concentration.

[0009] In this range, the measured concentration can be determined with the highest accuracy achievable by the method. Thanks to the high signal-to-noise ratio of the measurements, various self-checking and self-correction processes can be performed in this range, for example, in the case of the photoacoustic method the wavelength and / or the modulation frequency of the light source can be optimised with the help of processes caried out in this range, and thereby the accuracy and reliability of the measurements can be improved significantly.

[0010] As a rule of thumb, it can be ascertained that in the case of concentration measurements, more particularly, for example, in the case of photoacoustic measurements, the preferred S / N range is a partial range of the range 10<S / N<100,000. In addition to the preferred range, it is also worth addressing the case of S / N values higher and lower than the preferred values. The range of higher- than-preferred S / N values is in a way a saturation range that is typically characterised by a nonlinear relationship between the signal and the measured concentration, and in this range the noise typically also becomes concentrationdependent. It is thus not expedient to use this range for performing concentration measurements. Likewise, it is also not expedient to use the range of lower-than- preferred S / N values for concentration measurements, because in this range concentration measurements can only be carried out with high inaccuracy. In this range, in addition, systems performing measurements based on the photoacoustic principle typically have the following disadvantages:

[0011] Typically, there is a nonlinear relationship between the signal and the concentration. This phenomenon is related to the lock-in detection applied in photoacoustic measurements, as a result of which the photoacoustic signal behaves as a complex quantity, i.e., in addition to having a magnitude (amplitude) it also has a phase (here, phase can be interpreted as the time delay of the periodic microphone signal with respect to modulation). Nonlinearity is caused by the phase of the useful signal generated by the light absorption of the component to be measured being different from the phase of the so-called background signal generated by light absorption on the wall and / or the window of the chamber. In this case, the useful signal and the background signal are added as two complex numbers; and when complex numbers with unequal phase are added there is a nonlinear relationship between the magnitude of the resulting signal and the magnitude of the terms, i.e., in our case, the useful signal, that is, the concentration of the component to be measured. At low signal levels the noise can be higher than it is in the preferred S / N range; the reason for this is also that the signal and the background signal are out of phase.

[0012] Self-checking and self-correction processes cannot be carried out in this range, because the measured signals are too noisy for accurately determining the maximum of the absorption line of the component being measured and the maximum of the resonance curve of the chamber.

[0013] As can be seen from the comparison above, concentration measurements, for example, more specifically, photoacoustic measurements must be expediently performed in the range of preferred S / N values. However, in many cases where concentration-measurement instruments are applied the concentration of the component to be measured is low, so the generated signal is also low, and thus in most cases the measurement instruments do not operate in the preferred S / N range.

[0014] Attaining the preferred S / N is usually not an easy undertaking; various solutions can be tested for achieving the preferred S / N range using a concentration-measurement system based on the photoacoustic (or an alternative) operating principle. In the photoacoustic case, for example such lasers can be utilised - typically in the midinfrared range - that can be tuned to one of the strongest lines of the component to be measured. Unfortunately, however, these lasers are typically much more expensive and are also much larger and more complex than the so-called telecommunications lasers working in the near-infrared range, so - primarily for reasons of cost-efficiency - the application of the latter type is customary, even as the signal-to-noise ratio of the measurements performed using these lasers is not necessarily sufficient for performing self-checking (self-correction) processes.

[0015] In the context of the above, the concept of the sensitivity of the system can be introduced. It holds true in general that the measured signal is dependent on the concentration; in certain cases / ranges it is also true that: signal = sensitivity x concentration. If the sensitivity is adjustable, then with the help of adjusting it a preferred or optimal signal-to-noise ratio range can be reached; meanwhile of course it must be borne in mind, if necessary, that the method utilised for increasing sensitivity may also change the noise level of the measurement. In certain cases, for example, the low signal level corresponding to low concentrations, that is, a low signal-to-noise ratio can be increased by increasing the sensitivity, or, for example an overly high signal that generates a signal-to-noise ratio that is higher than the optimal can be brought into the preferred or optimal signal-to-noise ratio range by reducing the sensitivity.

[0016] With instruments utilising tunable-wavelength lasers and having photoacoustic or other optical operating principles, if the measurement is performed at a given absorption line of the component to be measured, and for example at the currently measured concentration the signal-to-noise ratio of the measurement is lower than optimal, then, in case the component to be measured has an absorption line in the tuning range of the laser where light absorption is greater than at the line used otherwise, by retuning the laser to this line the measured signal intensity and thereby the signal-to-noise ratio of the measurement can be increased, and desirably can be brought into the preferred or the optimal signal-to-noise ratio range. The reverse of this is also true, i.e. , in case a measurement has too high a signal-to-noise ratio at a given concentration value, the signal-to-noise ratio can be reduced by tuning to an absorption line where the signal is lower. However, tuning is on the one hand time-consuming, which implies that during the tuning operation the instrument is “blind”, i.e., it is not measuring the current concentration, and on the other hand by default the laser is usually tuned to the strongest line available in the tuning range, so it is impossible to tune it to an even stronger line. Tuning to a weaker line in the case of overly high concentrations also has the above-described disadvantage of time loss.

[0017] Sensitivity can also be increased by modifying the instrument to enable it to measure lower-than-usual concentrations. In the case of photoacoustic instruments or instruments with other optical principles of operation, this can also be achieved by applying a different (usually much more expensive) laser that can be tuned to a stronger absorption line. In the case of alternative methods, increasing sensitivity can only be achieved by including sensors that are significantly more expensive than usual sensors.

[0018] In addition to the S / N ratio, “focusing” the concentration measurement to a preferred range within the overall measurement range may be desirable due to other circumstances, parameters, and considerations. Furthermore, it is also desirable to improve the usability and sensitivity of existing concentration measurement systems without there being a need to replace the entire system, or certain system components with more expensive or more complex system components. Therefore, such a general solution is needed that is able to fulfil these objectives easily and effectively.

[0019] A prior art approach to increasing measured concentration under controlled circumstances and, as a result, operating a concentration measurement system based on the photoacoustic or another principle in a preferred S / N range is based on admixing to the gas to be measured a calibration gas, i.e. , a gas (additive gas) containing a known concentration of the gas component to be measured. If the concentration of the gas component to be measured is sufficiently high in the calibration gas, and the volume flow rate of the calibration gas is also set to an appropriate value, then the concentration of the gas component to be measured will already be sufficiently high in the gas mixture to provide that the concentration measurement system operates in a preferred S / N range. Furthermore, if the volume flow rates of both the original gas and the admixed calibration gas are known exactly, then the concentration of the gas component to be measured in the original gas to be measured can be calculated from the concentration measured in the gas mixture.

[0020] Solutions utilising additive calibration gas are disclosed for example in US 5,214,952, US 5,540,077, and JP7159323 A2. However, this known approach is less well suited for practical application, because the operation of calculating the original concentration from the measured concentration requires that both the original and the admixed calibration gas volume flow rates are known accurately. In many cases accurately measuring the volume flow rate of the gas to be measured is a particularly difficult task, because for instance the gas flow can be heavily contaminated, which makes accurate volume flow rate measurement practically impossible. The inaccuracy of volume flow rate measurement directly deteriorates the accuracy of concentration measurements, so this simple admixing-based method cannot be used for the above-described purposes, i.e. , for bringing the concentration measurement into a preferred range, for example in order to increase the signal-to-noise ratio.

[0021] In addition to the above, by “focusing” the concentration measurement to a preferred range within the overall measurement range in relation to the gas or liquid component being examined may necessitate not only a concentration increase but also concentration reduction (i.e., dilution), so a need persists for providing a universal solution.

[0022] DISCLOSURE OF THE INVENTION

[0023] The object of the invention is therefore to provide a universal concentration measuring apparatus and method that are able to simply and effectively determine the concentration of a component to be measured in a fluid flow such that the actual measurements are always performed in a preferred partial measurement range or preferred partial measurement ranges within an overall measurement range.

[0024] The object according to the invention has been achieved by the concentration measuring apparatus according to claim 1 and the concentration measuring method according to claim 11. Preferred embodiments of the invention are defined in the dependent claims.

[0025] In contrast to the prior art approaches, the apparatus and method according to the invention is suited for increasing or decreasing the concentration to be measured in a controlled manner in order to keep the signal-to-noise ratio or other parameters or characteristics of the measurements in a preferred range, and thereby for determining the concentration of the given component with high accuracy. A particularly preferable feature of the invention is that it achieves improved-accuracy concentration measurement applying existing concentration measurement devices, without necessitating the replacement of the system or certain system components with more expensive, more complex system components.

[0026] BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Preferred embodiments of the invention will be explained referring to the accompanying drawings, where

[0028] Fig. 1 is a schematic structural view of the apparatus according to the invention, Fig. 2 is a diagram showing preferred partial measurement ranges within the overall measurement range,

[0029] Fig. 3 is a table containing data of a measurement example, and

[0030] Fig. 4 is a diagram showing the relative deviation percentage in the example according to Fig 3 as a function of the split ratio of the additive fluid.

[0031] MODES FOR CARRYING OUT THE INVENTION

[0032] According to the basic idea of the apparatus and method according to the invention the original fluid flow, i.e. , liquid or gas flow, is split into two at a known volume flow rate split ratio; this can be done for example by applying, in the simplest way, equal splitting, and admixing to both resulting partial flows a respective partial flow of the calibration fluid such that the volume flow rates of the partial fluid flows of the calibration fluid are different but their ratio is known and constant. By measuring a concentration of the component to be measured in both mixed fluids, provided that the volume flow rate ratios are known in both of the split fluid flows in the course of splitting the original fluid flow, and the volume flow rate ratios of the two split fluid flows relative to the admixed calibration fluid are also known, the concentration in the original fluid of the component to be measured can be calculated mathematically

[0033] - in the case of a linear characteristic curve, by applying a simple mathematical formula, and in the case of nonlinear characteristics, applying a nonlinear function

[0034] - from the concentrations measured in both detector units. It is preferable if the above-mentioned mathematical calculation yields a concentration calculation formula that contains only volume flow rate split ratios instead of specific volume flow rates, because in this way the calculated concentration will only depend on the split ratios that can be determined accurately, i.e., it will not depend on volume flow rate values that can be very difficult or impossible to measure in actual practice. Because it is a much simpler task to keep the volume flow rate ratio of the partial fluid flows at a constant value than the volume flow rates themselves, with these embodiments concentration measurement with high accuracy and (for example) with a sufficiently high signal-to-noise ratio can be achieved especially effectively and easily.

[0035] The exemplary concentration measuring apparatus according to the invention illustrated in Fig. 1 is adapted for measuring a concentration Y of a component to be measured in a fluid flow, and comprises a feeding unit and detector units connected to the feeding unit that will be described in detail below.

[0036] The apparatus has a first detector unit 11 and a second detector unit 12, and the feeding unit has a first fluid splitter unit 21 that has an inlet configured to receive an examined volume flow of the fluid flow and a first and a second outlet configured to respectively output, in a first split ratio AE, a first partial volume flow X1and a second partial volume flow X2. The first outlet of the first fluid splitter unit 21 is in fluid communication with the first detector unit 11 , and the second outlet of the first fluid splitter unit 21 is in fluid communication with the second detector unit 12. The feeding unit further comprises a second fluid splitter unit 22 having an inlet configured to receive a volume flow of an additive fluid containing a known concentration c of the component and a first and a second outlet configured to respectively output, in a second split ratio AA, a first partial volume flow F1and a second partial volume flow F2different from the first partial volume flow F1. The first outlet of the second fluid splitter unit 22 is in fluid communication with the first detector unit 11 , and the second outlet of the second fluid splitter unit 22 is in fluid communication with the second detector unit 12. In the illustrated preferred embodiment the first outlet of the first fluid splitter unit 21 and the first outlet of the second fluid splitter unit 22 are connected to a first common conduit section 13 leading to the first detector unit 11 , expediently by way of conduits in a direct or indirect manner. Similarly, the second outlet of the first fluid splitter unit

[0037] 21 and the second outlet of the second fluid splitter unit 22 are directly or indirectly connected to a second common conduit section 14 leading to the second detector unit 12.

[0038] The exemplary apparatus further comprises a sampling unit 31 that is in fluid communication with the inlet of the first fluid splitter unit 21 and is configured to separate the examined volume flow from a conduit 30 carrying a fluid flow containing an unknown concentration of the component to be measured. This latter is applied for taking samples from the fluid flow to be measured.

[0039] For verifying the measurement results and for calibration it is preferable if the apparatus comprises volume flow rate measurement units 41 , 42, 43 adapted for respectively measuring volume flow rates through the first common conduit section 13, the second common conduit section 14, and the fluid flow conduit 30.

[0040] The calculations can be simplified if the first fluid splitter unit 21 has a first and a second outlet adapted for outputting, respectively, a first partial volume flow X1and a second partial volume flow X2, which partial volume flows being identical to each other. In the simplest case, the fluid control means responsible for splitting the unkown-concentration mixture are implemented as two identical outlet stubs through which the two partial fluid flows flow at an identical volume flow rate. Alternatively, the solution described below in relation to the second fluid splitter unit

[0041] 22 can also be applied.

[0042] However, the mixture must typically be split with the second fluid splitter unit 22 into volume flows that significantly differ from each other, so the unit preferably comprises a mass flow controller, a rotameter, or critical flow orifices. The most important characteristic of the fluid control means responsible for splitting the flow of the additive fluid (that has a known concentration) is that they always allow for the accurate and stable adjustment of the ratio of the volume flow rates in the two branches in a repeatable manner.

[0043] In the preferred embodiment according to Fig. 1 , the first detector unit 11 and the second detector unit 12 are configured as respective photoacoustic chambers having identical measurement characteristic curves and expediently identical structure, and the system further comprises a laser light source 50 that is provided with control and processing electronics, is configured for photoacoustic concentration measurement, and operates at the wavelength of the specific measurement. The preferred embodiment therefore comprises two photoacoustic chambers adapted for receiving respective gas mixtures with different concentrations of the gas component to be measured. Each of the photoacoustic chambers has a gas outlet port 15, 16.

[0044] The apparatus preferably further comprises a container 32 for the additive fluid, said container 32 being in controllable fluid communication with the inlet of the second fluid splitter unit 22. Preferably, a conduit extends between the container 32 and the inlet of the second fluid splitter unit 22, with a controllable shut-off valve or a changeover valve configured to direct a flow of the additive fluid through a unit adapted for absorbing the component. The additive (or calibration) fluid contains the component to be measured in a known concentration that is sufficiently high in case of concentration augmentation and is sufficiently low or zero in case of concentration diminution. That is, the concentration c of the component to be measured in the additive fluid may be higher or lower with respect to the concentration Y of the component in the fluid flow; in order to provide a larger margin of adjustability, it is expediently significantly lower or higher than that, for example at least by a factor of ten, by a factor of a hundred, or by a factor of a thousand, or is lower by the same factor.

[0045] The invention can be supplemented with additional units and functionalities that further improve reliability. For example, in case the composition of the major components of the calibration fluid is significantly different from the composition of the fluid to be measured, and, as a result, one or more physical or chemical properties of the fluid to be measured and the calibration fluid significantly differ, then, by measuring these properties separately in the fluid flows entering the two measuring chambers the volume flow rate ratio of the fluid flows can be determined, and thereby the measurement can be verified. One of these possibly measurable properties is sonic speed, i.e. , in case the sonic speed is different in the fluid to be measured and the calibration fluid due to the composition of the gases or liquids, the mixture rate can be determined from the sonic speed measured in the mixed fluid, and the quality of admixing can be checked. Thus, a respective sonic speed measurement unit configured to measure sonic speed in the flowing fluid is arranged in each of the first common conduit section 13 and the second common conduit section 14.

[0046] The concentration measuring method according to the invention is therefore adapted for measuring a concentration Y of a component to be measured in a fluid flow; the measurement can be performed applying the concentration measuring apparatus described above. Proper operation is ensured such that the apparatus has a first detector unit 11 and a second detector unit 12 that have respective measurement characteristic curves comprising a characteristic curve section that is identical (has identical shape) in both measurement characteristic curves; this is indispensable for the improvement in accuracy achieved by simple calculation based on “duplicated” operation. This common characteristic curve section, or any partial range within it can be applied for defining a partial measurement range within the overall measurement range that makes possible such measurements that are preferable in some respect.

[0047] The measurement range corresponding to this characteristic curve section comprises a first partial measurement range CR1provided for the first detector unit 11 and a second partial measurement range CR2provided for the second detector unit 12 that does not overlap with the first partial measurement range CR1.

[0048] The method comprises feeding an examined volume flow of the fluid flow to the inlet of the first fluid splitter unit 21 , and feeding a first partial volume flow X1and a second partial volume flow X2split into two by the first fluid splitter unit 21 in a first split ratio AE into the first detector unit 11 and the second detector unit 12, respectively, and further comprises feeding a volume flow of an additive fluid containing a known concentration c of the component to the inlet of the second fluid splitter unit 22, and feeding a first partial volume flow Fi and a second partial volume flow F2different from the first partial volume flow F1, split into two in a second split ratio AE, into the first detector unit 11 and the second detector unit 12, respectively,

[0049] In the course of method, the partial volume flows X1, X2, F1, F2are chosen such that when a concentration measurement of the component is performed with the first detector unit 11 , a first concentration value C1falling into the first partial measurement range CR1is obtained, and when a concentration measurement is performed with the second detector unit 12, a second concentration value C2falling into the second partial measurement range CR2is obtained. These adjustments can be performed for example by trial-and-error, but the calculations detailed below may also help to achieve measurement results falling into the appropriate ranges.

[0050] Applying the method steps described above, two measurement values falling onto the common measurement characteristic curve are obtained; more specifically, knowing the common characteristic curve section, the concentration Y of the component can be calculated based on the first measured concentration value C1and the second measured concentration value C2.

[0051] Using the results of the concentration measurements performed on the fluids fed to the two detector units 11 , 12, for example, gases fed to the measurement chambers, the concentration of the component to be measured in the original fluid can be preferably calculated in such a way that we do not have to know the volume flows in the individual fluid carrying branches, i.e., only the volume flow ratios characteristic of the splitting of the original fluid and the calibration fluid are required. For the case of a linear characteristic curve section, this can be proven by performing the following calculation:

[0052] Y is the concentration of the component to be measured in the original fluid flow, c is the concentration of the component to be measured in the additive calibration fluid flow, C1and C2are the respective concentrations of the component to be measured in the first and second detector units 11 , 12, X1is the partial volume flow of the original fluid in the first detector unit 11 ,

[0053] AE is the ratio of the partial volume flows of the original fluid in the second and first detector units 12, 11 , X2= AE-X1

[0054] FI is the partial volume flow of the additive fluid in the first detector unit 11 ,

[0055] AA is the ratio of the partial volume flows of the additive fluid in the second and first detector units 12, 11 , F2= AA-F1

[0056] Rearranging the above: and from (3): by transforming (3): The obtained formula proves the above statement, i.e. , the preferable properties of the linear case of the invention, because the concentration Y of the component to be measured in the original fluid flow can be calculated based on the measured concentration values C1and C2, the concentration c of the measured component in the calibration fluid, and the volume flow split ratios AE and AA.

[0057] It is therefore expedient to select a linear characteristic curve section for the measurement, because in such a case the concentration Y of the component can be calculated easily using the following formula, without knowing the specific volume flow values: where

[0058] AE is the first split ratio,

[0059] AA is the second split ratio,

[0060] C is the concentration of the component in the additive fluid, C1is the first measured concentration value, and C2is the second measured concentration value.

[0061] If the characteristic curve section constitutes a nonlinear function, the concentration Y of the component is calculated based on a nonlinear function of the characteristic curve section fitted on the first C1measured concentration value and on the second C2measured concentration value. Also in this case, it is expedient to select a characteristic curve section from which such a formula can be obtained that only necessitates the knowledge of the split ratios without there being a need to know the specific volume flow rate values.

[0062] Measurement accuracy of the method according to the invention primarily stems from the fact that the measurement is focused on a preferable partial measurement range. It is, however, preferable if the two measured values are as far apart from each other in this partial range as possible, i.e. they are as close to the boundaries of the partial range as possible. This is because the measurement according to the invention essentially boils down to fitting a straight line or a curve to specific values. Instead of the single measurement point applied conventionally, we generate two measurement points, and determine the concentration from the slope of the line or the path of the curve fitted on the points. The further apart these two points are from each other, the more accurate the fitting, i.e. , our ability to determine concentration. It is preferable, therefore, if the first partial measurement range CR1is the bottom third, preferably the bottom 25%, more preferably the bottom 10%, of the measurement range corresponding to the characteristic curve section, and the second partial measurement range CR2is the upper third, preferably the upper 25%, more preferably the upper 10%, of the measurement range corresponding to the characteristic curve section.

[0063] The diagram depicted in Fig. 2 shows the signal S and the signal-to-noise ratio S / N as a function of concentration. The figure shows linear function curve shapes, but specific measurements can of course also yield nonlinear function curves. In the figure, the area with inclined hatching indicates the measurement range characterizable with the saturation / nonlinear behaviour mentioned in the introduction, i.e., it is not expedient to perform measurements in a concentration range above a signal-to-noise ratio value S / Nmax. The vertically hatched area indicates a region where the disadvantages related to the lower S / N values (also mentioned in the introduction) occur, so it is also expedient to avoid regions under the value S / Nmin. Therefore, measurements must stay within the measurement range between Cmin and Cmax defined by the values S / Nmin and S / Nmax. It is preferable to omit the range with horizontal hatching because measured concentration values C1and C2that are close to each other reduce the accuracy of the measurement. It is therefore expedient to choose the first partial measurement range CR1and the second partial measurement range CR2for example according to Fig. 2, and ensure that the measured concentration values C1and C2are close to the lower and upper boundaries thereof.

[0064] The characteristic curve section and the measurement range corresponding thereto are therefore preferably chosen to be between a lower limit of a signal-to-noise characteristic curve of the measurement determined based on a noise condition and an upper limit thereof determined based on a saturation condition.

[0065] It can be set forth as a general rule that the preferred S / N range is typically a central portion (partial range) of the overall measurement range, i.e., a range of concentrations that are not overly high and not overly low. This implies that the limits of this central portion should typically be determined empirically for a given measurement, which usually only means an approximation. The bottom of the preferred S / N range can be for example at one-third or one-fifth of the measurement range, while the top can be for example at 2 / 3 or 4 / 5 thereof. If, for example, the top and the bottom of the measurement range are 300 ppm and 0.5 ppm, then the exemplary preferred S / N ranges can be the following: 100-200 ppm, 60-240 ppm, but even 60-200 ppm or 100-240 ppm. Here, the S / N value relates to the signal produced by a given concentration, i.e., the relationship between the bottom of the preferred S / N range and the measured concentration is that the corresponding measured concentration produces a signal having a magnitude for which the S / N value (magnitude) will already reach the bottom of the preferred S / N range.

[0066] The S / N characteristics of the detector units 11 , 12 can be determined in a known manner. In case the invention is used based on the S / N characteristics, it is imperative to know the S / N characteristics well for optimal measurement effectiveness. Therefore, during the initial calibration of the detector units 11 , 12 not only the calibration curve (i.e., the relationship between the measured signal and the concentration) must be determined, but also how the S / N value varies as a function of the measured signal. For example, in the field of photoacoustics, S / N characteristics typically undergo significant changes with changing conditions. Knowing the specific S / N characteristics, on the one hand the split ratio to be applied can be determined, and on the other hand a calculated value up to which it is worth applying the invention and the value at which it is worthwhile to stop applying the invention (i.e., to return to the conventional way of measurement) can also be determined in advance. If the concentration to be measured increases drastically, causing the measurement signal to fall into the preferred S / N range even without admixing the additive fluid, then the method according to the invention does not need to be applied because in such a case the measurement can be performed using a single detector unit at an appropriate S / N value. In such a case it can be worthwhile to assess whether the accuracy of concentration measurement deteriorates if we nevertheless use the method according to the invention. If there is no significant deterioration, then it is still preferable to apply the method according to the invention. If, however, the concentration to be measured is so high that the application of the method would deteriorate measurement accuracy, then it is expedient to stop the method for the time the concentration to be measured is high, and return to the basic operation of the analytic instrument, i.e., to performing the measurement applying a single detector unit. The occurrences of overly high concentration can be easily detected applying the invention, i.e., after stopping the system and waiting for the measured concentration to fall under the limit value, restarting the system can be implemented easily. For example, the system can be stopped applying a shut-off valve which stops the injection of the additive fluid, i.e., Fi and F2are reduced to zero; a changeover valve can also be included in the system, and, if necessary, the additive fluid can be directed through a chemical substance that absorbs the additive component. In such a case the concentration calculation formula must be modified by inserting zero as the value of the concentration of the additional fluid.

[0067] If, therefore, the calculated concentration Y of the component falls into the measurement range corresponding to the characteristic curve section, i.e., it would fall into the preferred range without the apparatus and operation according to the invention, then, preferably as a verification or for further improving accuracy, the concentration Y of the component is measured with the first detector unit 11 and / or the second detector unit 12 also without the addition of an additive fluid.

[0068] Fig. 3 shows the results of an exemplary measurement performed applying an apparatus and method according to the invention. According to the example, a photoacoustic hydrogen sulphide measurement used in the natural gas industry is performed with the following parameters: Component to be measured: hydrogen sulphide (H2S)

[0069] Calibration range of the photoacoustic chambers: 5 - 100 ppm

[0070] Concentration of the test sample (Y): 5 ppm

[0071] Error of the single-detector (i.e., conventional) measurement: 1 ppm

[0072] Gas to be admixed (c): hydrogen sulphide at 1000 ppm in an ammonia gas mixture Base flow rate (X1=X2; AE=1 ): 100 cm3 / min = 0.1 l / min

[0073] The 1 ppm error of the single-detector measurement means that with a single detector the relative error of measuring a concentration of 5 ppm is 20%. As can be seen from the table, if AA is chosen to be 1.5, then the error of the measurement applying two detector units will exceed 20%, so with these settings it is not worthwhile using the tow-detector (dual-injection) method according to the invention. If the value of AA is chosen to be three, the solution according to the invention already yields a relative error under 8%, so it already proves to be more favourable than the conventional measurement. By further increasing AA, the accuracy of the two-detector measurement can be improved further.

[0074] Fig. 4 is a diagram showing the relative difference, expressed as a percentage, of the example according to Fig 3, i.e., the relative error of dY / Y as a function of the split ratio of the additive fluid; as can be seen in the diagram, the curve drops sharply at the start and then flats out.

[0075] In most cases, the expected average value of the component concentration Y is known, for example from prior experience or from preliminary approximate measurements. In such cases, the partial volume flows X1, X2, F1, F2are preferably chosen based on the expected average value of the component concentration Y, for example according to the following:

[0076] The first partial volume flow X1and the second partial volume flow X2of the examined volume flow of the fluid flow are chosen to be of identical magnitude, the first partial volume flow F1of the additive fluid being calculated applying the formula where X = X1= X2are the flow rates of the partial volume flows that have been chosen to be identical, cn is the measured concentration to be achieved in the first partial measurement range CR1,

[0077] Yais the expected average value of the component concentration, and c is the concentration of the component in the additive fluid, while the second partial volume flow F2is calculated applying the formula where

[0078] X = X1= X2are the flow rates of the partial volume flows that have been chosen to be identical, ct2 is the measured concentration to be achieved in the second partial measurement range CR2,

[0079] Yais the expected average value of the component concentration, and c is the concentration of the component in the additive fluid.

[0080] In the above-described manner the desired measurement values, and therefore the desired measurement accuracy can be preferably achieved.

[0081] The aim of the technical solution according to the invention is to increase / decrease the measured concentration values by a given extent such that the signal-to-noise ratio or other parameter of the measurements performed at the measured concentration values falls into the preferred range. This optimisation can be implemented by the appropriate selection of the volume flow ratios (first and foremost the split ratio AA) and the concentration c of the component to be measured in the calibration fluid. Furthermore, if the fluid to be measured contains a component that also generates a signal, i.e. , a so-called “cross-effect” occurs, then it is expedient to optimise the method also in a way that the concentration of the component to be measured is sufficiently high in the admixed calibration fluid, because that way the cross-effect can be reduced or even eliminated.

[0082] In addition to increasing / adjusting the signal-to-noise ratio, the invention may also be suited for reducing or eliminating cross-effects in case the concentration of the component to be measured in the calibration fluid is sufficiently high. On the other hand, it is also possible to check the operation of the measurement chambers, i.e. , the detector units 11 , 12, for example by removing (applying a so-called “decontaminant”) the component to be measured from the fluid flow to be measured, or by applying a “zero fluid”, for example “zero gas” instead of the fluid to be measured. In such a case, knowing the concentration of the component to be measured in the calibration fluid and the volume flow split ratios AE and AA, it can be known precisely what concentrations should be measured in the two detector units 11 , 12, for example, photoacoustic measurement chambers; i.e., it can be checked whether in both measurement chambers those concentrations are measured which should be measured based on the calculations. In such a way a self-check procedure of the system can be performed.

[0083] The invention may also allow for checking if the current split ratios AA, AE are the same as the split ratios that were originally set based on which the concentration Y of the component to be measured is calculated. If it is known for certain that during a given measurement period the concentration Y to be measured is constant, for instance because it is kept constant artificially, or the characteristics of the process itself being examined are such that the concentration Y to be measured is constant over a given time period, then the following process can be applied:

[0084] In a conventional manner, we vary the concentration c of the component to be measured in the admixed additive fluid, and with the help of the formula related to the concentration Y, using the initially set split ratio values AA and AE and the measured concentrations C1and C2, we calculate the values of Y for the various c values. If the AA and AE values used for calculating Y are actually identical to the originally set values of AA and AE, then the Y values calculated from the C1and C2values measured at different concentrations c will be constant values. However, if the current AA and AE values are different from the initial values of AA and AE that are applied in the formulas, the concentration Y calculated from the concentrations C1and C2measured with different c values will not be constant but will vary as a function of c. In the latter case it can be concluded that the originally set AA and / or AE value has been changed. Based on the results of the self-check procedure performed as set forth above, the error caused by the varying values of AA and / or AE can be corrected, either by restoring the original AA and AE values or by using the changed AA and AE values, so the accuracy of concentration values measured and calculated using the invention can be restored.

[0085] The solution according to the invention can be advantageously applied in all such cases where the concentration of the component to be measured is so low that the produced signal barely exceeds measurement noise, i.e., the signal-to-noise ratio of the measurement is low. In relation to the fluid to be measured there are no restrictions on the application of the method, but the method can be applied particularly preferably for measuring heavily contaminated gases or liquids, for example exhaust gases, natural gas, and gases and liquids produced in complex industrial processes.

[0086] The invention can also be applied for determining the concentrations (Y) of more than one components by applying an additive fluid that contains all components in known concentrations (c). Such a solution is equivalent to multiplying the method according to the invention.

[0087] Multiplicated systems are used also for critical concentration measurement applications, for example in the natural gas industry, where for increased safety typically two instruments are used for concentration measurement instead of using only one, the two instruments checking each other’s operation. The present invention also allows for using two apparatuses at the same time, in which case the split ratios AA and AE can be chosen in many combinations, i.e., they can be identical or different in the two systems. In a particularly preferred embodiment, additive fluids (for example calibration gas bottles) with different concentrations c are used for the two apparatuses. In this embodiment, for the concentrations Y to be measured that both apparatuses are able to measure well, the application of two apparatuses makes cross-verification possible; on the other hand, the measurement range of the method according to the invention can also be extended, because such concentration values Y may exist at which one of the apparatuses is able to perform good or better-quality measurements, for example by applying lower concentrations of the additive fluid c, while at other values the other apparatus is able to provide good or better-quality measurements. Of course, even more than two apparatuses can be applied, providing additional possibilities for cross-verification and extending the measurement range.

[0088] The apparatus and method according to the invention can be applied not only in photoacoustic cases but also for any type of concentration measurement of gases or liquids. It holds true in general that an analytical (concentration-measurement) instrument has a measurement range, i.e., a range between the lowest and the highest measurable concentration values; this measurement range has the characteristic that for low concentrations the measurement is noisy, i.e., the measurement has a poor signal-to-noise ratio; however, by increasing the concentration to be measured, the measurement becomes less noisy, that is, the signal-to-noise ratio of the measurement improves. At high concentrations, problems related to saturation typically occur, and in any analytical concentration measurement it may be preferable to keep the focus of the measurement on a selected measurement range also for reasons other than the signal-to-noise ratio. Applying two detector units and the injection method according to the invention, and using the appropriate formulas for the calculations, therefore, the concentration measurements performed in any analytic method can be transferred to some specific measurement range, and thus the accuracy of the measurement can be improved.

Claims

CLAIMS1 . A concentration measuring apparatus for measuring a concentration (Y) of a component to be measured in a fluid flow, comprising a feeding unit and a detector unit coupled to the feeding unit, characterised by having a first detector unit (11) and a second detector unit (12), the feeding unit comprising a first fluid splitter unit (21) having an inlet configured to receive an examined volume flow of the fluid flow and a first and a second outlet configured to respectively output, in a first split ratio (AE), a first partial volume flow (X1) and a second partial volume flow (X2), wherein the first outlet of the first fluid splitter unit (21) is in fluid communication with the first detector unit (11), and the second outlet of the first fluid splitter unit (21) is in fluid communication with the second detector unit (12), and a second fluid splitter unit (22) having an inlet configured to receive a volume flow of an additive fluid containing a known concentration (c) of the component and a first and a second outlet configured to respectively output, in a second split ratio (AA), a first partial volume flow (F1) and a second partial volume flow (F2) different from the first partial volume flow (F1). wherein the first outlet of the second fluid splitter unit (22) is in fluid communication with the first detector unit (11), and the second outlet of the second fluid splitter unit (22) is in fluid communication with the second detector unit (12).

2. The apparatus according to claim 1 , characterised in that the first outlet of the first fluid splitter unit (21) and the first outlet of the second fluid splitter unit (22) are directly or indirectly connected to a first common conduit section (13) leading to the first detector unit (11), and the second outlet of the first fluid splitter unit (21) and the second outlet of the second fluid splitter unit (22) are directly or indirectly connected to a second common conduit section (14) leading to the second detector unit (12).

3. The apparatus according to claim 1 or 2, characterised by comprising a sampling unit (31 ) that is in fluid communication with the inlet of the first fluid splitter unit (21) and is configured to separate the examined volume flow from a conduit (30) of the fluid flow.

4. The apparatus according to any of claims 1 to 3, characterised by comprising volume flow rate measurement units (41 , 42, 43) respectively measuring volume flow rates through the first common conduit section (13), the second common conduit section (14), and the fluid flow conduit (30).

5. The apparatus according to any of claims 1 to 4, characterised in that the first and the second outlets of the first fluid splitter unit (21) are adapted to output identical first and second partial volume flows (X1, X2).

6. The apparatus according to any of claims 1 to 5, characterised in that the second fluid splitter unit (22) comprises a mass flow controller, a rotameter, or critical flow orifices.

7. The apparatus according to any of claims 1 to 6, characterised in that the first detector unit (11) and the second detector unit (12) are configured as respective photoacoustic chambers having identical measurement characteristic curves, and the system further comprises a laser light source (50) provided with control and processing electronics configured for photoacoustic concentration measurement.

8. The apparatus according to any of claims 1 to 7, characterised by comprising a container (32) for the additive fluid, said container (32) being in controllable fluid communication with the inlet of the second fluid splitter unit (22).

9. The apparatus according to claim 8, characterised by comprising a conduit between the container (32) and the inlet of the second fluid splitter unit (22), with a controllable shut-off valve or a changeover valve configured to direct a flow of the additive fluid through a unit adapted for absorbing the component.

10. The apparatus according to claim 2, characterised in that a respective sonic speed measurement unit configured to measure sonic speed in the flowing fluid is arranged in each of the first common conduit section (13) and the second common conduit section (14).

11. A concentration measuring method for measuring a concentration (Y) of a component to be measured in a fluid flow, characterised by applying for the measurement the concentration measuring apparatus according to claim 1 with a first detector unit (11) and second detector unit (12), measurement characteristic curves of which comprise a characteristic curve section that is identical in both measurement characteristic curves, wherein a measurement range corresponding to said characteristic curve section comprises a first partial measurement range (CR1) provided for the first detector unit (11) and a second partial measurement range (CR2) provided for the second detector unit (12) that does not overlap with the first partial measurement range (CR1), the method comprising feeding an examined volume flow of the fluid flow to the inlet of the first fluid splitter unit (21), and feeding a first partial volume flow (X1) and a second partial volume flow ( X2) split into two in a first split ratio (AE) by the first fluid splitter unit (21) into the first detector unit (11) and the second detector unit (12), respectively, feeding a volume flow of an additive fluid containing a known concentration (c) of the component to the inlet of the second fluid splitter unit (22), and feeding a first partial volume flow (F1) and a second partial volume flow (F2) different from the first partial volume flow (F1), split into two in a second split ratio (AE), into the first detector unit (11) and the second detector unit (12), respectively, wherein the partial volume flows (X1, X2, F1, F2) are chosen such that when a concentration measurement of the component is performed with the first detector unit (11), a first concentration value (C1) falling into the first partial measurement range (CR1) is obtained, and when a concentration measurement is performed with the second detector unit (12), a second concentration value (C2) falling into the second partial measurement range (CR2) is obtained, and calculating the concentration (Y) of the component based on the first measured concentration value (C1) and on the second measured concentration value (C2).

12. The method according to claim 11 , characterised in that the characteristic curve section is linear, and the concentration (Y) of the component is calculated applying the formulawhereAE is the first split ratio,AA is the second split ratio, c is the concentration of the component in the additive fluid, C1is the first measured concentration value, and C2is the second measured concentration value.

13. The method according to claim 11 , characterised in that the characteristic curve section constitutes a nonlinear function, and the concentration (Y) of the component is calculated based on a nonlinear function of the characteristic curve section fitted to the first measured concentration value (C1) and the second measured concentration value (C2).

14. The method according to claim 11 , characterised in that the first partial measurement range (CR1) is the bottom third, preferably the bottom 25%, more preferably the bottom 10%, of the measurement range corresponding to the characteristic curve section, and the second partial measurement range (CR2) is the upper third, preferably the upper 25%, more preferably the upper 10%, of the measurement range corresponding to the characteristic curve section.

15. The method according to claim 11 , characterised in that the partial volume flows (X1, X2, F1, F2) are selected based on an expected average value of the component concentration (Y).

16. The method according to claim 15, characterised in that the first partial volume flow (X1) and the second partial volume flow (X2) of the examined volume flow of the fluid flow are chosen to be of identical magnitude, and the first partial volume flow ( F1) of the volume flow of the additive fluid is calculated applying the formulawhereX = X1= X2are the flow rates of the partial volume flows that have been chosen to be identical, cn is the measured concentration to be achieved in the first partial measurement range (CP1),Yais the expected average value of the component concentration, and c is the concentration of the component in the additive fluid, and the second partial volume flow (F2) is calculated applying the formulawhereX = X1= X2are the flow rates of the partial volume flows that have been chosen to be identical, ct2 is the measured concentration to be achieved in the second partial measurement range (CP2),Yais the expected average value of the component concentration, and c is the concentration of the component in the additive fluid.

17. The method according to claim 11 , characterised in that the characteristic curve section and the measurement range corresponding thereto are chosen to be between a lower limit of a signal-to-noise characteristic curve of the measurement determined based on a noise condition and an upper limit thereof determined based on a saturation condition.

18. The method according to claim 11 , characterised in that the concentration (c) of the component in the additive fluid is higher than the concentration (Y) of the component to be measured in the fluid flow.

19. The method according to claim 11 , characterised by measuring, in case the calculated concentration (Y) of the component falls into the measurement range corresponding to the characteristic curve section, the concentration (Y) of the component with the first detector unit (11) and / or the second detector unit (12) also without the addition of an additive fluid.

20. The method according to claim 11 , characterised by determining the concentrations (Y) of more than one components by applying an additive fluid that contains all components in known concentrations (c).

Images