JP2016156836A - Method and apparatus for siloxane measurements in biogas - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method of monitoring one or more silicon-containing compounds present in a biogas, in particular siloxane and trimethyl-silanol.SOLUTION: A method includes generating a first absorption spectrum on the basis of a ratio of a first spectral measurement and a second spectral measurement, where the first spectral measurement is obtained from a non-absorptive gas having substantially no infrared absorptions in a specified wavelength range of interest, and the second spectral measurement is obtained from a sample gas comprising the biogas. The method includes generating at least one surrogate-absorption spectrum on the basis of individual absorption spectrum for each of silicon-containing compounds in a subset of one or more silicon-containing compounds selected from a larger set of known silicon-containing compounds with known concentrations. A total concentration of the one or more silicon-containing compounds in the biogas can be calculated on the basis of the first absorption spectrum and the at least one surrogate absorption spectrum.SELECTED DRAWING: None

Description

  The present invention relates generally to absorption spectrometers. More particularly, it relates to monitoring and measuring the concentration of siloxane compounds, for example in biofuels or biogas. This technique also relates to determining at least one of the total concentration of all siloxane compounds or the total concentration of all silicon-containing compounds in a biofuel or biogas or the like.

  Spectroscopy is a method of examining the interaction between electromagnetic radiation and a sample (eg, including one or more of a gas, a solid, or a liquid). How radiation interacts with a particular sample depends on the properties of the sample (eg, molecular composition). In general, when radiation passes through a sample, a particular wavelength of radiation is absorbed by molecules within the sample. The particular wavelength of radiation that is absorbed is unique to each molecule within a particular sample. Thus, by identifying which wavelengths of radiation are absorbed, it is possible to identify specific molecules present in the sample.

Infrared spectroscopy is a spectroscopic specific method that, for example, determines the type of molecule and the concentration of individual molecules in the sample by applying infrared electromagnetic energy to the sample (eg, gas, solid, liquid, or combination thereof). Is a field. In general, infrared energy is characterized as electromagnetic energy having a wavelength of energy between about 0.7 [mu] m (frequency 14000 cm ^-1) of about 1000 .mu.m (frequency 10 cm ^-1). Infrared energy travels through the sample and the energy interacts with molecules in the sample. The energy that passes through the sample is detected by a detector (eg, an electromagnetic detector). The detected signal is then used, for example, to determine the molecular composition of the sample and the concentration of a particular molecule within the sample.

  One particular type of infrared spectrometer is a Fourier Transform Infrared (FTIR) spectrometer. Such infrared spectrometers are used in various industries such as air quality monitoring, explosive / biological weapon detection, semiconductor processing, and pharmaceutical manufacturing. Different applications of FTIR spectrometers require different sensitivities to allow the user to identify which molecules are present in the sample and to determine the concentration of the various molecules. Depending on the application, it may be necessary to specify the concentration of individual molecules in a sample with an accuracy of about 1 billion parts per billion (ppb). In industrial applications, it is necessary to increase the sensitivity, so by optimizing existing spectroscopic systems and using new spectroscopic components, the system repeatedly breaks down the concentration of molecules in the sample. And it becomes possible to carry out reliably.

An FTIR spectrometer may be used, for example, to monitor the concentration of a compound in a gas. Biofuel (eg, biogas) is used to drive a variety of equipment including turbine generators. Biogas burns and drives the equipment. Biogas (eg, animal waste, wastewater, or gas from landfills) may include a variety of compounds including siloxane compounds. The siloxane compounds in the biogas also burn to produce oxides (eg, SiO ₂ (eg, silica or sand)). SiO ₂ adheres to the turbine blades and also to the turbine bearings, thus reducing turbine performance or possibly causing the turbine to fail. This deposition process accelerates as the level of siloxane in the biogas increases. Biogas producers typically use activated carbon filters to capture siloxanes, but siloxane levels increase when the filters are depleted.

Conventional methods for monitoring the concentration of siloxane compounds in biogas are performed off-line by analyzing samples taken from biogas. For example, conventional techniques comprise separating and measuring siloxane from background gas using GC / MS (ie, gas chromatography / mass spectroscopy) techniques. To analyze the sample gas, the sample is grab-sampled for analysis and run on a GC / MS system. Typically, a field sample is taken from the gas stream and introduced into a stainless steel canister or Tedlar ™ sample bag or collected using a methanol solvent impinger. The sample is then transferred to an analysis laboratory for analysis. It usually takes a few days to understand the analysis results.

  The sample tends to condense the sample components, thus making it difficult to evaluate the true composition of the sample. Samples so removed may only have time to analyze the contents once and therefore may not represent the true composition in the sample. GC / MS analysis of a sample can take several hours to analyze the siloxane compound in the sample and can take too long to allow the operator to intervene. If the siloxane level rises, you may have already missed the opportunity to implement usable measures. Increasing the ability to monitor and measure the concentration of siloxanes and / or silicon-containing materials in biogas can extend the life of the turbine. Furthermore, if siloxane and / or silicon-containing materials can be monitored, detected and quantified quickly, a longer time can be secured for applying available means / interventions.

Spectroscopy can be used, for example, to detect, identify, and / or quantify trace siloxanes and other silicon-containing compounds in biogas (eg, sample biogas with an accuracy of about 1/5 billion ppb) The concentration of the individual siloxane compounds in it may be specified). Trace amounts of cyclic siloxanes in biogas (eg D3-siloxane, D4-siloxane, D5-siloxane, D6-siloxane), linear siloxanes (eg L2-siloxane, L3-siloxane, L4-siloxane, L5-siloxane) And trimethylsilanol (TMS) may be detected and quantified. The concentration of siloxane or silicon can generally be measured in-situ in real time, such as landfills, animal excretion sites, or wastewater pools (eg, taking a sample locally and taking the sample for a considerable amount of time). Process and analyze the contents of the sample biogas without the need to analyze in the laboratory after that). In-line continuous monitoring can detect siloxane and / or total silicon level increases in real time and notify the operator or automatically stop the process to prevent unnecessary exposure of the turbine to SiO _2. .

Samples that contain substances (eg, compounds) (eg, coherent absorbers) that have substantially higher infrared absorptance compared to other substances in the sample depend on the FTIR applying infrared energy to the sample. Thus, problems may arise with FTIR spectroscopy. Interfering absorbers in the sample prevent the concentration of other substances detected in the sample having a substantially lower infrared absorption rate from being effectively detected and measured in the sample. Biogas may include molecules such as siloxanes and other silicon-containing compounds, hydrocarbon compounds (eg, methane or ethane), water, or carbon dioxide. Hydrocarbon compounds in biogas are relatively high infrared absorption at specific wavelengths compared to siloxane compounds (eg, D4 siloxane has an absorptance of about 0.001 at a wavelength of about 7.8 microns). (E.g., when the wavelength is methane, the absorption is about 0.055 at about 7.8 microns). Thus, the hydrocarbon can be an interfering absorber. Siloxane compounds may have relatively high infrared absorption at wavelengths ranging from about 8 microns to about 12 microns (eg, D4 siloxane has an absorption of about 0.075 at about 8.2 microns). The absorption rate is 0.125 at about 11 microns. Thus, the concentration of the siloxane compound in the sample biogas is a spectral measurement in the wavelength range of interest (eg, about 8 microns to about 12 microns), even in the presence of hydrocarbon compounds or other interfering absorbers. It may be measured by obtaining. As the target wavelength range, a wavelength range in which the absorbance of the main components (for example, H ₂ O, CO ₂ , CH ₄ ) of the biogas is not high may be selected. Siloxanes and TMS compounds may have absorbances that overlap with other hydrocarbons in the wavelength range of interest. Multivariate analysis methods may be used to distinguish the contributions of siloxane compounds and other hydrocarbons and to strictly evaluate the contributions of siloxane compounds and / or silicon-containing compounds.

  In one aspect, a method is provided for monitoring one or more silicon-containing compounds present in a biogas, the method comprising supplying a non-absorbing gas to a sample cell. Non-absorbing gases do not substantially absorb infrared in the specified wavelength range of interest. The method includes obtaining a first spectral measurement from the sample cell and supplying biogas to the sample cell. The method further includes obtaining a second spectrum measurement from the sample cell and generating a first absorption spectrum based on a ratio of the first spectrum measurement and the second spectrum measurement. The method includes at least one based on an individual absorption spectrum for each silicon-containing compound of a subset of one or more silicon-containing compounds selected from a larger set of known silicon-containing compounds having a known concentration. Generating two surrogate absorption spectra. The total concentration of the one or more silicon-containing compounds in the biogas may be calculated based on the first absorption spectrum and the at least one surrogate absorption spectrum.

  In another aspect, a computer readable product is provided that is tangibly embodied on a non-transitory information carrier or machine readable storage device and operable on a digital signal processor for a biogas detection system. The computer readable product includes instructions operable to cause the digital signal processor to receive a first spectral measurement of the non-absorbing gas in the sample cell. Non-absorbing gases do not substantially absorb infrared radiation in the specified wavelength range of interest. Further, the computer readable product includes instructions operable to cause the digital signal processor to receive a second spectral measurement of the sample gas comprising the biogas in the sample cell. The computer readable product causes a digital signal processor to generate a first absorption spectrum based on a ratio of the first spectral measurement to the second spectral measurement, and at least a known silicon-containing compound having a known concentration. And further comprising instructions operable to generate a set of surrogate absorption spectra based on the individual absorption spectra for each silicon-containing compound of the subset of one or more silicon-containing compounds selected from the larger set. The computer readable product causes a digital signal processor to perform multiple regression analysis using the set of first absorption spectra and surrogate absorption spectra to calculate the total concentration of one or more silicon-containing compounds in the biogas. Further comprising operable instructions.

  In other examples, any of the above aspects may include one or more of the following features. In some embodiments, a correlation coefficient is applied to the total concentration. The correlation coefficient scales the total concentration by some factor. In some embodiments, the one or more silicon-containing compounds in the biogas comprises at least one siloxane. Larger collections of known silicon-containing compounds may also include at least one siloxane. The subset of one or more silicon-containing compounds may also include at least one siloxane. In some embodiments, the total concentration is of the total concentration of siloxane compounds in the biogas, the total concentration of other silicon-containing compounds in the biogas, or the total concentration of all silicon-containing compounds in the biogas. It consists of one.

In some embodiments, the subset of one or more silicon-containing compounds is selected based on a spectral match between the known silicon-containing compound and one or more silicon-containing compounds present in the biogas. . In some embodiments, at least one silicon-containing compound of the subset of one or more silicon-containing compounds is present in the biogas. In some embodiments, at least one silicon-containing compound of the subset of one or more silicon-containing compounds is not present in the biogas.

  In some embodiments, a larger collection of known silicon-containing compounds is D3-siloxane, D4-siloxane, D5-siloxane, D6-siloxane, L2-siloxane, L3-siloxane, L4-siloxane, and L5-siloxane. Consists of. The subset of one or more silicon-containing compounds may consist of 3 to 5 silicon-containing compounds selected from a larger set of known silicon-containing compounds.

  In some embodiments, the biogas comprises landfill gas. In this case, the subset of one or more silicon-containing compounds is a) L2-siloxane, L3-siloxane, and D4-siloxane, b) L2-siloxane, D3-siloxane, and D4-siloxane, or c) L2 It may consist of one of siloxane, D3-siloxane, and D5 siloxane.

  In some embodiments, the biogas consists of digested biogas. In this case, a subset of the one or more silicon-containing compounds is a) D3-siloxane, D5-siloxane, and L3-siloxane, b) D4-siloxane, D5-siloxane, and L3-siloxane, or c) D3. -Consisting of one of siloxane, D5-siloxane, and L2 siloxane.

  In some embodiments, the surrogate absorption spectrum is an individual absorption for each hydrocarbon compound of a subset of one or more hydrocarbon compounds selected from a larger set of known hydrocarbon compounds having a known concentration. Further includes a spectrum. A larger collection of known hydrocarbon compounds can consist of ethane, propane, and butane. The surrogate absorption spectrum includes at least an individual absorption spectrum for each silicon-containing compound in a subset of one or more silicon-containing compounds and an individual absorption spectrum for each hydrocarbon compound in a subset of one or more hydrocarbon compounds. It may be a model based on.

  In some embodiments, calculating the total concentration of the one or more silicon-containing compounds in the biogas is performed using a processor and a multiple regression analysis using the first absorption spectrum and the surrogate absorption spectrum. Including performing. Multiple regression analysis may be performed using classical least squares (CLS), partial least squares (PLS), inverse least squares (ILS), or principal component analysis (PCA). The total concentration value may be determined such that the surrogate absorption spectrum is substantially similar to the first absorption spectrum. In some embodiments, the total concentration of one or more silicon-containing compounds in the biogas may be calculated locally in real time.

In some embodiments, the second spectral measurement may be obtained over an acquisition period of about 10 seconds to about 20 seconds.
In some embodiments, biogas is obtained from animal waste, wastewater, or landfill.

In another aspect, a system for monitoring one or more silicon-containing compounds in biogas is provided. The system includes a light source of a first radiation beam, an interferometer that receives the first radiation beam from the light source and forms a second radiation beam of interference signals, and a sample cell in optical communication with the interferometer. Including. The system further includes a flow mechanism that establishes a first flow of non-absorbing gas that does not substantially absorb infrared radiation in the specified wavelength range of interest, and a second flow of biogas through the sample cell. . The system includes a cooling detector in optical communication with the sample cell. The cooling detector is adapted to receive a first interference signal propagating in the non-absorbing gas in the sample cell and a second interference signal propagating in the sample gas in the sample cell. The system further includes a processor in electrical connection with the cooling detector. The processor is selected from 1) a first absorption spectrum based on a ratio of the first interference signal and the second interference signal, and 2) at least a larger set of known silicon-containing compounds having a known concentration. Calculate the total concentration of one or more silicon-containing compounds in the biogas based on a set of surrogate absorption spectra based on the individual absorption spectra for each silicon-containing compound in the subset of one or more silicon-containing compounds It is configured as follows. The system further includes a housing in which the light source, interferometer, sample cell, cooling detector, and processor are disposed.

  In some embodiments, the sample cell of the system includes a concave reflective field mirror surface at the first end of the sample cell. The sample cell may further include a substantially spherical concave reflective objective mirror surface positioned in opposing relationship with the field mirror surface at the second end of the sample cell. The objective surface matches each focal point in at least one plane so as to maximize the throughput of the second radiation beam propagating in the sample cell via multiple reflections at each of the field mirror surface and the objective surface. It has a cylindrical component that increases the degree.

  In some embodiments, the set of surrogate absorption spectra is at least an individual absorption spectrum for each silicon-containing compound and a known concentration of a known hydrocarbon compound having a known concentration of one or more silicon-containing compound subsets. A model based on individual absorption spectra for each hydrocarbon compound in a subset of one or more hydrocarbon compounds selected from a large set.

Other aspects and advantages of the invention will become apparent from the following drawings, detailed description, and claims, all illustrating the principles of the invention by way of example only.
The advantages of the invention described above, together with further advantages, may be better understood by reference to the following description in conjunction with the accompanying drawings. In the drawings, like reference characters generally refer to the same parts throughout the different views. The drawings are not necessarily to scale, emphasizing overall the principles of the invention.

1 is a block diagram of an exemplary detection system for monitoring and / or detecting trace gases in a gas sample according to the present invention. 1 is a schematic diagram of an exemplary optical configuration according to the present invention. 1 is a block diagram of an exemplary flow system for introducing a sample into a sample cell according to the present invention. 6 is a graph comparing the optical path length / NEA and the number of passes between the optical surfaces of the sample cell according to the present invention. FIG. 3 is a graph comparing trace gas concentration versus time when trace gas is input to an exemplary detection system according to the present invention. FIG. The figure which shows the timeline of a series of measured values by this invention. 1 is a plan view of an exemplary detection device for monitoring and / or detecting trace gases in a gas sample according to the present invention. FIG. FIG. 3 is a plan view of several components of an exemplary detection device for monitoring and / or detecting trace gases in a gas sample according to the present invention. 2 is a flowchart illustrating a method for monitoring siloxane compounds in biogas, according to an illustrative embodiment of the invention. FIG. 4 is a diagram representing a NIPALS decomposition of spectral information represented by matrix X (spectrum measurements) and matrix Y (concentration data), according to an illustrative embodiment of the invention. FIG. 4 shows individual absorption spectra used to monitor silicon-containing compounds in biogas, according to an exemplary embodiment of the present invention. 6 is a flowchart illustrating another exemplary method for monitoring silicon-containing compounds in a biogas sample. The figure which shows the result of the total siloxane density | concentration with time regarding a synthetic landfill gas sample. The figure which shows the result of the total siloxane density | concentration with time of a digestion gas sample.

  FIG. 1 shows a block diagram of an exemplary apparatus 10 for monitoring and / or detecting trace gases in a gas sample. Device 10 may be used to detect trace substances such as sarin, tabun, soman, sulfur mustard, and VX nerve gas. The device 10 may be used, for example, to detect the level of siloxane in biogas. In some embodiments, vapors of solid or liquid substances may be detected. The apparatus 10 may be an absorption spectrometer and / or a Fourier transform infrared (FTIR) spectrometer. In the exemplary embodiment, apparatus 10 includes light source 14, interferometer 18, sample cell 22, source of gas sample 26, detector 30, processor 34, display 38, and housing 42. Including. In various embodiments, the device 10 may be used to detect a trace amount of gas in a short time with little or no false positives or false negatives.

In various embodiments, the light source 14 may generate a radiation beam (eg, an infrared radiation beam). The light source 14 may be a laser or an incoherent light source. In one embodiment, the light source is a glow bar, ie, an inert solid that is heated to about 1000 ° C. to produce black body radiation. The glow bar may be formed from silicon carbide and may be energized. Spectral range of the system may be between about 600 cm ^-1 to about 5000 cm ^-1. The resolution of the system may be 2 cm ⁻¹ and about 4 cm ⁻¹ . In one embodiment, the detection system may record a higher resolution spectrum of the trace gas upon detection of the trace gas. Higher resolution spectra can help identify trace gases.

  In various embodiments, the radiation source 14 and the interferometer 18 may comprise a single instrument. In some embodiments, interferometer 18 is a Michelson interferometer that is widely known in the art. In one embodiment, interferometer 18 is a BRIK interferometer commercially available from MKS Instruments, Inc. (Wilmington, Mass., USA). The BRIK interferometer is a combiner that splits and combines incident radiation, a movable corner cube to modulate the radiation, a white light source used to identify the central burst, and a vertical to monitor the corner cube velocity. And a cavity surface emitting laser (VCSEL). The BRIK interferometer is not affected by tilt and lateral motion errors, and is not affected by thermal fluctuations, thus increasing the robustness of the interferometer.

In one embodiment, interferometer 18 may be a module that includes a radiation source, a fixed mirror, a movable mirror, an optical module, and a detector module (eg, detector 30). The interferometer module can measure all optical frequencies generated by its light source and transmitted through the sample (eg, sample 26 contained within sample cell 22). The radiation is directed to an optical module (eg, a beam splitter), which can split the radiation into two beams, a first signal and a second signal. The movable mirror initially forms a variable optical path length difference between two substantially identical beams of electromagnetic energy. The movable mirror is usually moved or moved at a constant speed. After the first signal has propagated a different distance than the second signal (in this embodiment, due to the movement of the movable mirror), the optical module recombines the first signal and the second signal. A radiometric signal having a modulated intensity can be generated by the interference of the two beams. This interference signal passes through the sample and is measured by a detector. The presence of various samples (eg, solid, liquid, or gas) can modulate the intensity of the radiation as detected by the detector. Thus, the detector output becomes a variable time dependent signal depending on the optical path difference established by the relative position of the fixed and movable mirrors and the modulation of the electromagnetic signal generated by the sample. This output signal can be represented as an interferogram.

  The interferogram can be represented as a plot of the received energy intensity versus the position of the movable mirror. Those skilled in the art refer to interferograms as signals that are a function of time. The interferogram is a function of the variable optical path difference generated by the displacement of the movable mirror. The interferogram is referred to as a “time domain” signal by those skilled in the art because it typically moves the position of the movable mirror at a constant rate. An interferogram may be understood as the sum of all wavelengths of energy emitted by a light source and passing through a sample. A computer or processor can use a Fourier transform (FT) mathematical method to convert the interferogram to a spectrum that characterizes light absorbed through or transmitted through the sample. Since each type of molecule absorbs a specific wavelength of energy, it is possible to determine the molecules present in the sample based on the interferogram and the corresponding spectrum. Similarly, the amount of energy absorbed by or transmitted through the sample may be used to determine the concentration of molecules in the sample.

  In various embodiments, the interference signal is formed without using an interferometer. An optical signal is recorded using an absorption spectrometer and information about trace species is derived from the signal transmitted through the sampling area. For example, an absorption spectrum or a difference spectrum may be used.

  In various embodiments, the sample cell 22 may be a bent optical path and / or a multi-pass absorption cell. The sample cell 22 may include an aluminum housing that encloses the system of optical components. In some embodiments, the sample cell 22 is a bent path optical analysis gas cell as described in US Pat. No. 5,440,143, the disclosure of which is incorporated herein.

  In various embodiments, the source of the sample of gas 26 may be the atmosphere. The sample cell 22 or gas sampling system may collect ambient air and introduce it into the sampling region of the sample cell 22. A sample of gas may be introduced into the sample cell 22 at a predetermined flow rate using a flow system that includes an inlet 46 and an outlet 50 of the sample cell 22.

In various embodiments, detector 30 may be an infrared detector. In some embodiments, detector 30 is a cold detector. For example, the detector 30 may be a cold cooled detector (eg, a cadmium mercury telluride (MCT) detector), a Stirling ™ cooled detector, or a Peltier cooled detector. In one embodiment, the detector is a deuterated triglycine sulfate (DTGS) detector. In one embodiment, the detector is a 0.5 mm Stirling cooled MCT detector with a 16 μm notch that can provide the sensitivity necessary to detect trace gases. The relative responsivity (i.e. the ratio of responsivity as a function of wavelength) of the Stirling cooled MCT detector is at least 80% over the dominant wavelength region of interest (e.g. 8.3 [mu] m to 12.5 [mu] m). Furthermore, the D ^* value of the Stirling cooled MCT detector may be at least 3 × ^{10 10} cmHz ^1/2 W ⁻¹ . D ^* may be defined as the inverse of the detector equivalent noise power multiplied by the square of the active element area.

The processor 34 may receive the signal from the detector 30 and identify the trace gas by a trace fingerprint of the trace gas or indicate the relative or absolute concentration of a particular substance in the sample. The processor 34 may be, for example, signal processing hardware and quantification analysis software running on a personal computer. The processor 34 may include a processing unit and / or memory. The processor 34 may acquire and process the spectrum continuously while calculating the concentrations of a plurality of gases in the sample. The processor 34 may send information such as trace gas identification, trace gas spectrum, and / or trace gas concentration to the display 38. The processor 34 may store the spectral concentration time history in a graph and tabular format and may store the measured spectrum and spectral residuals, which may be displayed. The processor 34 may collect and store various other data for later reprocessing or review. Display 38 may be a cathode ray tube display, a light emitting diode (LED) display, a flat screen display, or other suitable display known in the art.

  In various embodiments, the housing 42 may be configured to provide a detection system having one or more of portability, robustness, and light weight. The housing 42 may include a handle and / or be easily secured to a transport mechanism such as a pull cart or hand truck. The housing 42 may be strong enough to prevent the optical system from shifting or components from being destroyed when transported and / or dropped. In various embodiments, the device 10 may weigh on the order of 40 pounds (18.14 kg). In one embodiment, the device 10 is fully self-contained (eg, collecting all the components in the housing 42 necessary to collect the sample, record the spectrum, process the spectrum, and display information about the sample). Including).

  FIG. 2 illustrates an exemplary embodiment of an optical configuration that can be used in the apparatus 10. Radiation from the light source 14 (eg, Glover) is directed by the first mirror 52 to the interferometer 18 (eg, including a potassium bromide beam splitter). The radiation beam is directed by the parabolic mirror 54 (PM) to the first folding mirror 58 and into the sample cell 22. The radiation beam exits the sample cell and is directed by the second folding mirror 62 to the elliptical mirror 66 (EM), which directs the radiation beam to the detector 30.

  In one exemplary embodiment, the parabolic mirror 54 has an effective focal length of about 105.0 mm, an apparent focal length of about 89.62 mm, and may have an eccentricity value of about 74.2 mm. . The diameter of the parabolic mirror 54 may be about 30.0 mm, and the reflection angle may be about 45 °.

  In one embodiment, elliptical mirror 66 may have a major radius of about 112.5, a minor radius of about 56.09, and an ellipse tilt angle of about 7.11 °. The diameter of the elliptical mirror 66 may be about 30.0 mm, and the reflection angle (chief ray) may be about 75 °.

In various embodiments, the first folding mirror 58 is about 25 mm in diameter and the second folding mirror 62 is about 30 mm in diameter.
The mirror and optics may include a gold coating, a silver coating, or an aluminum coating. In one embodiment, the elliptical and parabolic mirrors are coated with gold and the flat folding mirror is coated with silver.

In various embodiments, the sample cell may include an objective mirror surface 74 and a field mirror surface 78. The objective mirror surface 74 may be substantially spherical and concave. The field mirror surface 78 may be concave and may be located in a relationship facing the objective mirror surface 74. The objective surface 74 may include at least one cylindrical component that increases the coincidence of each focal point in at least one plane so as to maximize the throughput of the radiation beam propagating between the surface 74 and the surface 78. In one embodiment, the objective surface 74 may include a plurality of substantially spherical concave reflective objective surfaces, each surface being in focus in at least one plane to maximize the throughput of the radiation beam. Cylindrical components that enhance the degree may be included. The center of curvature of the objective mirror surface may be located behind the field mirror surface 78. By increasing the focus coincidence in at least one plane, distortion, astigmatism, spherical aberration, and coma can be better controlled and higher throughput can be achieved. The addition of cylindrical components can reduce the effective radius of curvature in one plane, thus allowing light incident on the reflective surface to better approach the focal point in the orthogonal plane. In one embodiment, cylindrical components are superimposed on the objective mirror surface 74 to form various radii of curvature in two orthogonal planes. The objective surface 74 may have a contour that approximates a toroidal shape.

The total optical path length of the sample cell 22 may be between about 5 m and about 15 m. However, longer optical path lengths and shorter optical path lengths may be used depending on the application. In one detailed embodiment, the sample cell 22 has a total optical path length of about 10.18 m due to the total number of passes between the objective mirror surface 74 and the field mirror surface 78 being about 48 times. The optics of the sample cell 22 may be optimized to obtain a 0.5 mm detector and a 1 steradian collection angle. The detector optics magnification ratio may be about 8: 1. Objective mirror surface 74 and field mirror surface 78 may have a gold coating between about 800 cm ^{−1 and} 1200 cm ⁻¹ with a nominal reflectivity of about 98.5%. The internal volume of the sample cell may be between about 0.2L and about 0.8L. However, a larger volume and a smaller volume may be used depending on the application. In one detailed embodiment, the volume is about 0.45 L.

  In one embodiment, the radiation beam is incident on the sample cell 22 and is directed through the sample cell 22 to be focused on the radiation beam on the entrance slit of the sample cell 22 and / or the radiation beam is directed toward the detector. The mirrors and optics used in the process may be optimized to match the optical properties of the sample cell, thereby maximizing radiation throughput and increasing the sensitivity of the detection system.

For example, in one embodiment, a properly aligned optical configuration can be about 88.8% efficient. As used herein, efficiency may be the ratio of the number of rays incident on the image to the total number of rays emitted within the emission angle range. In one embodiment, the positions of the folding mirrors 58, 62 as well as the detector 30 may be adjustable, and thus various between the interferometer 18, the parabolic mirror 54, the sample cell 22 and the detector 30. It becomes possible to compensate for mechanical tolerance errors. In one embodiment, the following nominal (design) optical distance may be used to optimize throughput.
The distance (X1) from the detector to the elliptical mirror is about 21.39 mm.
・ The distance (X2) from the elliptical mirror to the folding mirror is about 132.86 mm.
・ The distance (X3) from the folding mirror to the sample cell (the surface of the field mirror) is about 70.00 mm.
The sample cell optical path length is about 10181.93 mm.
-The distance (X4) from the sample cell to the folding mirror is about 70 mm.
-The distance (X5) from the folding mirror to the parabolic mirror is about 35 mm.

  FIG. 3 illustrates an exemplary embodiment of an exemplary flow system 82 for introducing a sample into the sample cell 22. The flow system 82 includes a filter 86, a flow sensor 90, an optional heating element 90, a gas cell 22, a pressure sensor 98, a valve 102, and a pump 106 coupled to the gas piping 110. Arrows indicate the direction of flow. One or more of the components of the flow system 82 may include, for example, Teflon ™, to withstand decontamination temperatures, resist CWA and TIC corrosivity, and avoid condensing siloxanes. Stainless steel and wettable parts such as Kalretz ™ may be included.

Filter 86 may be an in-line 2 μm stainless steel filter commercially available from Mott Corporation (Farmington, Conn.). The flow sensor 90 may be a mass flow sensor that includes stainless steel wettability parts, such as a flow sensor commercially available from McMillan Company (Georgetown, TX). The heating element 94 may be a line heater commercially available from Watlow Electric Manufacturing Company (St. Louis, MT). The pressure sensor 98 may be a Baratron ™ pressure sensor commercially available from MKS Instruments (Wilmington, Mass.). Valve 102 may be stainless steel, may include a Teflon ™ O-ring, and may be a valve commercially available from Swagelok (Solon, Ohio). . The gas pipe 110 may be a pipe having a diameter of 3/8 "(9.5 mm) commercially available from Swagelok.

  The pump 106 may be a “micro” diaphragm pump with a heated head. A Dia-Vac ™ B161 pump commercially available from Air Dimensions, Inc. (Deerfield Beach, Florida) may be used. In one embodiment, a small diaphragm pump commercially available from Hargraves Technology Corporation (Mooresville, NC) may be used. In the exemplary embodiment, pump 106 may be located downstream of sample cell 22 to draw air into sample cell 22. In some embodiments, a pump without a heated head may be used. In some embodiments, a pump is not required if the gas sample is not pressurized. In such a situation, the upper pressure of the gas sample causes the sample to propel through the gas cell. As a result, any leaks in the system can be pulled away from the analyzer rather than pushed into the analyzer, minimizing the possibility of contaminating the internal components of the analyzer. Furthermore, unwanted products of unintended chemical reactions to the pump elastomer can be prevented from entering the sample cell 22.

  In various embodiments, the flow rate through the flow system 82 may be between 2 L / min and 10 L / min. However, a higher flow rate and a lower flow rate may be used depending on the application. In one embodiment, the flow rate is between 3 L / min and 6 L / min. The sample pressure may be about 1 atmosphere. However, a higher pressure and a lower pressure may be maintained depending on the application. In some embodiments, the sample cell may be operated at a high pressure, such as up to 4 atmospheres. The operating temperature of the sample cell may be between about 10 ° C and about 40 ° C. However, higher and lower temperatures may be maintained depending on the application. In one embodiment, the detection system may include a heating element for heating the sample between about 40 ° C and about 180 ° C. In one embodiment, the temperature may be raised to about 150 ° C. to remove contamination of the device.

In various embodiments, the sample cell optical path length may be between about 5 m and about 12 m. The spacing between the field mirror surface and the objective surface may be limited by the gas sampling flow rate. In one embodiment, a 5.11 m sample cell with a spacing of 16 cm and a pass count of 32 may have an internal volume of about 0.2 L. In another embodiment, if the number of passes is the same, the volume can be about 0.4 L when the number of passes is 32 and the spacing is 20.3 cm. In yet another embodiment, if the spacing is 25.4 cm, the volume can be about 0.6L. A flow rate that can supply an appropriate amount of “fresh” ambient gas at least every 10 seconds may be determined. However, a smaller sampling rate may be realized. In various embodiments, the flow rate (eg, between 2 L / min and 10 L / min) may be optimized to achieve an optimal gas exchange rate. For example, in one embodiment, the gas exchange rate is at least 80% when the detection time interval is 20 seconds. In one embodiment, the gas exchange rate is about 80% when the detection time interval is 10 seconds.
To about 95%.

The optical path length / NEA ratio may be used as a metric for quantifying the sensitivity of the detection system. The optical path length is the total beam optical path length of the sample cell measured in meters, and NEA is the noise equivalent absorbance measured in absorbance units (AU). If the sensitivity is limited by the chance error of the detection system (also called random noise, detector and electronics noise), the detection limit can be inversely proportional to the optical path length / NEA ratio. For example, if this ratio is doubled, the detection limit for a particular sample in units of ppb or mg / cm ^{3 is} halved. This is therefore an appropriate quantification metric for sensitivity performance. This metering does not take into account sensitivity improvements due to advanced sampling techniques such as gas pressurization and cooling trapping.

Considering system noise limitations such as detector noise and digitization noise, the optical path length / NEA ratio can be optimized for various system configurations. Parameters that can be optimized include flow rate, sample cell volume, optical path length, number of passes in the sample cell, optical configuration, specular reflectivity, specular reflective material, and detector used. For example, the optimal detector is the detector that has the highest D ^* value and speed (shorter response time) within size, cost, and lifetime constraints.

For a spectrometer with limited detector noise, the sensitivity of the optical path length / NEA ratio is proportional to the D ^* value. The detector bandwidth can determine the maximum scan rate, which determines the maximum number of data averages that can be performed within an acceptable measurement period. In systems with limited detector or electronics noise, sensitivity generally increases with the number of averaged scans or, for example, the square root of the time to perform these scans. In one embodiment, for a Stirling cooling detector, the optical path length / NEA sensitivity ratio may be at least 1.5 × 10 ⁵ m / AU. DTGS detectors can be an inexpensive alternative because of their low cost and long maintenance-free life. However, the DTGS detector may have a lower D ^* value and be slower.

  The optical path length / NEA value may be determined by optimizing the distance between the field mirror surface and the objective mirror surface and the number of paths between these surfaces. FIG. 4 shows the path length / NEA as a function of mirror reflection for various surface spacings, eg, 6.3 inches (16.0 cm), 8 inches (20.3 cm), and 10 inches (25.4 cm). A graph is shown. As shown in FIG. 4, the maximum optical path length / NEA value occurs when the number of passes is about 92. However, when the number of passes is 92, only 25% of the light is transmitted due to reflection loss on the mirror surface. In one detailed embodiment, the sample cell has a transmission between about 50% and about 60%. When the mirror reflectivity is 98.5%, a transmittance of 60% corresponds to about 32 passes represented by a vertical line in FIG. A transmittance of 50% corresponds to about 48 passes. Table 1 shows exemplary combinations of parameters for realizing a sampling system for detecting trace gases in a sample.

  Table 1 Exemplary combinations of parameters that enable sampling of a trace gas detection system in a sample

The optical path length / NEA ratio may be converted to a detection limit in mg / m ³ or parts per billion (ppb) of concentration. The method used for such conversion is a comparison of expected peak absorbance and expected NEA value. Device 10 can be used to detect trace substances such as siloxane, sarin, tabun, soman, sulfur mustard, and VX nerve gas at concentrations below about 500 pp. In various embodiments, the concentration may be between about 10 ppb and about 500 ppb. However, higher and lower concentrations may be detected depending on the system and application. In some embodiments, the concentration may be between 5 ppb and about 50 ppb depending on the species. For example, the device 10 can detect traces of sarin at concentrations between about 8.6 ppb and about 30 ppb, and trace amounts of tabun at concentrations between about 12.9 ppb and about 39 ppb. Of soman can be detected at concentrations between about 7.3 ppb and about 22.8 ppb, trace amounts of sulfur mustard can be detected at concentrations between about 36.7 ppb and about 370.6 ppb, VX nerve gas can be detected at concentrations between about 12.9 ppb and about 43.9 ppb.

The gas exchange rate, i.e., the measured value of the accumulated amount of fresh gas supplied to the sample cell, combined with the optical path length / NEA ratio, "X mg / m ³ (or ppb) gas Y detected in Z seconds" The detection system response time specified as can be obtained. The detection system response time includes measurement time and calculation time (eg, about 5 seconds). Table 2 shows exemplary detection system response times for various substances such as sarin, tabun, soman, sulfa mustard, and VX nerve gas.

  Table 2 Exemplary detection system response times for trace gases measured using the detection system of the present invention (all response times are in seconds)

  FIG. 5 is a graph of trace gas concentration versus time using a step profile input (eg, trace gas enters the sample cell at the start of a measurement cycle). The measurement period “A” is the time during which data is collected and / or the interferogram is recorded. The calculation period “B” is when the interferogram is converted to a spectrum and spectral analysis is performed to generate data that allows determination of alarm levels and / or concentration values.

  FIG. 6 shows a timeline of a series of measured values. Substance 1 enters the sample cell and is detected during measurement period 1. During the calculation period 1, the interferogram is analyzed. During the measurement period 1, the substance 2 enters the sample cell. Substance 2 can be detected during the remainder of measurement period 1 if the concentration is sufficiently high. If substance 2 is undetectable, it is detected during a subsequent measurement period, eg, measurement period 2, and the interferogram is analyzed during a subsequent calculation period, eg, calculation period 2.

  In one embodiment, each reading can be separated in time at certain predetermined intervals. In various embodiments, this interval may be between about 1 second and about 1 minute. However, shorter intervals or longer intervals may be used depending on the application. In some embodiments, the interval is about 5 seconds, about 10 seconds, or about 20 seconds. Thus, the response time depends on this interval as well as when the substance becomes detectable by the detection system.

  In various embodiments, the detection system may detect trace gases, threat level, time of day, number of people in a room or building that may be affected by a substance, specific measurement application or scenario, or combinations thereof, etc. One or more parameters based on external factors may be employed. For example, in high threat conditions, shorter intervals may be used to minimize detection time and maximize trace substance detectability. In low threat situations, longer intervals that can maintain the lifetime of the detection system and reduce the possibility of false alarms (false positives or false negatives) may be used.

  Furthermore, individual measurements that exceed a threshold level for a particular substance can trigger the detection system to shorten the interval so that additional measurements can be made in a shorter time. In various embodiments, the first spectrum may be recorded with a first resolution or sensitivity. If contaminants are detected, the second spectrum may be recorded with a higher resolution or sensitivity, respectively. In addition, the detector may have a standby mode that operates at a higher temperature, thereby reducing the sensitivity of the detector. When triggered by an external factor, the temperature of the detector can be lowered to improve the sensitivity of the detector.

In various embodiments, the detection system can change the number of scans based on external factors or perceived threats. For example, the sensitivity of the detection system may be improved by increasing the number of scans. In one embodiment, the detection system can operate at a higher resolution while recording these additional scans. In one embodiment, the average number or the number of individual scans may be increased in each scan.

In various embodiments, the detection system only digitizes the low frequency region of the spectrum (eg, below 1300 cm ⁻¹ ) and can therefore scan faster. Electronic filter or detector response functions may be used to remove higher frequency regions (eg, greater than 1300 cm ⁻¹ ), thereby preventing or minimizing aliasing.

  In some embodiments, the detection system can detect the presence of trace gases in a portion of the spectrum. The second portion of the spectrum may be analyzed to confirm the presence of trace gas and / or to determine the trace gas concentration level.

  In one embodiment, the detection system can be packaged as a small self-contained multi-gas analyzer. For example, the detection system may be a diagnostic tool for recording, charting, analyzing and reporting air quality. 7 and 8 illustrate an exemplary detection system for monitoring air quality, for example ambient air of trace gases. As can be seen with reference to FIG. 7, the detection system is coupled to the housing 42 ′, the first display 38 ′, the second display 38 ″, the gas inlet 46 ′, the gas outlet 50 ′, and an external device. Port 118 to be connected.

  The housing 42 'may be a three-dimensional rectangular box that includes a top panel 122, a side panel 126, and a bottom panel 130 (shown in FIG. 8). The top panel 122 is removable from the side panel 126 by a hinge, so that the housing 42 'may be opened in use. The outer surface of the top panel 122 may include a first display 38 ′ and a second display 38 ″ attached to or embedded in the outer surface. The first display 38 ′ may be, for example, a touch It may be a liquid crystal display (LCD) having a screen display, the first display 38 ′ may receive commands for operating the detection system and may display a graphical user interface (GUI). The second display 38 "may be, for example, a light emitting diode (LED) display that includes a series of LEDs that are lit to indicate threat level, alarm status, and / or detection system status status. For example, the second display 38 ″ includes a first series of green, yellow, and red LEDs that indicate alarm status, and a second series of green, yellow, and red LEDs that indicate sensor status. In various embodiments, the housing 42 'may define a hole for taking in ambient air, which is used to introduce a sample of gas into the flow system to allow detection in the sample cell. be able to.

FIG. 8 shows an internal view of the top panel 122 and the bottom panel 130 when the top panel 122 is opened by a hinge. The bottom panel includes an inner body that includes an optics box 134 for housing optical components. The optics box 134 may be formed from an aluminum shell (eg, 6061-T6). In one embodiment, the optics box 134 is an airtight box. As shown in FIG. 8, the optical box 134 includes a light source 14 ', an interferometer 18', a sample cell 22 ', a detector 30', a parabolic mirror 54 ', and a first folding mirror 58'. And a second folding mirror 62 ', an elliptical mirror 66', an objective mirror surface 74 ', and a field mirror surface 78'. The optics box 134 may include a flow system that includes a valve 138 for regulating the gas flow rate, a pressure sensor 98 ′, a pump 106 ′, a gas line 110, and a fitting 142 for coupling. . A power supply 146 for the various components and fan 150 may be attached to the bottom panel 130. The detection system can be operated in still air and the fan 150 can maintain the internal temperature of the system. Bottom panel 13
0 also includes a connector 154 that interfaces with the top panel 122.

  As shown in FIG. 8, electronic components may be attached to the top panel 122. For example, the top panel 122 may include a data acquisition module 158, a mirror motion control module 162, a single board computer 166, a power distribution module 170, and a hard drive 172. The data acquisition module 158 may include a preamplifier, an analog to digital converter, and a data acquisition board. The preamplifier may amplify the analog signal received from detector 30 '. The analog signal may be converted to a digital signal using an analog to digital converter. The data acquisition board may be a Netburner processor board commercially available from Netburner (San Diego, Calif.). The single board computer 166 may be a readily available PC motherboard that runs Windows and presents a GUI to the user.

  The power distribution module 170 may process and distribute power to other modules in the system and may implement status sensors and status sensors that are used to monitor the functionality of the detection system. For example, the power distribution module 170 may distribute AC power to the system power supply 146 and the fan 150, and may include a temperature controller 174, such as Dwyer Instruments, Inc. [Michigan City, Ind. You may control Love Controls (trademark) commercially available from Michigan City). The power distribution module 170 also monitors the sample cell pressure, the differential pressure in the air filter, the sample cell temperature, and the detector temperature, A / D converts the output, and sends the result back to the single board computer 166. The power distribution module 170 may control the cooler of the Stirling cooling detector under commands from the single board computer 166. The top panel 122 may include a sample cell temperature transmitter.

  Data processing can be performed using modules that can be attached to the top panel 122 and allow real-time analysis of the data. The spectral library may include spectral fingerprints of between about 300 and about 400 gases. However, more gas may be added when recording spectra. Data processing can be done by standard computer programming languages such as MATLAB or C ++. The recorded spectrum may be transferred to MATLAB for post-spectral processing to calculate gas concentration, spectral residual, and / or false alarm rate. In various embodiments, the number of false alarms that occur during operation of the detection system is less than about 6 times per year. False alarms can be caused by noise, anomalous spectral effects, analysis codes, model errors, spectral library errors, or unknown interfering substances.

  The computer software may run on a Java-based platform with graphical remote control capabilities. The computer software may incorporate standard services including a user login, web-based GUI, alarm triggering, and / or an Ethernet ™ interface to a client computer that may be located remotely from the detection system. Good. The computer software may perform a remote health care diagnosis. In addition, the port 118 may be used to connect the system to a stand-alone computer capable of performing data processing and data analysis.

The housing 42 'is designed to withstand 50G impact. In one embodiment, the housing 42 ′ may be about 406 mm long and about 559 mm wide. The mass of the detection system may be about 20 kg. The housing 42 'may be attachable to a wall, a movable cart, or a wheelbarrow, and may include a handle (not shown) for carrying manually or using a mechanical lifting device. In one embodiment, the housing may be attached as part of a building air conditioning system. When the detector detects the presence of a contaminant, remedial action can be taken to account for that contaminant. For example, an alarm may sound and the building may be exhausted, or air flow within the air conditioning system may be increased to blow contaminants away from public areas or to dilute trace gases to an acceptable level.

  In various embodiments, the detection system may be operated at elevated temperatures to remove system contamination if contamination occurs. The system can heat the sample cell and flow system to between about 150 ° C. and about 200 ° C., while the remaining components, including electronics and optical components, are maintained at a temperature below about 70 ° C. You may be comprised so that. For example, components heated to about 150 ° can be isolated from surrounding components to prevent electronics damage and optical component readjustment or damage. When the sample cell and flow system operate at high temperatures, contaminant desorption can be accelerated. In one embodiment, the detection system may be operated while the system is decontaminated, and thus the progress of decontamination can be monitored. In one embodiment, the detection system is purged with nitrogen gas or ambient air during decontamination. The gas may contain moisture (eg, a relative humidity of about 30% or higher). In various embodiments, the system can remove contamination in less than about 2 hours and can be ready to return to a usable state.

  In one embodiment, the concentration of contaminant in the detection system may be determined, and if the concentration of contaminant exceeds a contamination value, at least the sample region may be heated to the decontamination temperature to remove the contaminant. Contaminant concentration may be monitored while heating the sample area, and heating may be reduced or terminated when the contaminant concentration reaches a decontamination value. The contamination value may be the concentration of a substance that inhibits the performance of the detection system. The decontamination value may be the concentration of a substance that can operate the detection system without being affected by the contaminant.

  In various embodiments, the sample cell of the detection system may be operated at high pressure. Although the optical path length / NEA ratio does not change, the sensitivity of the detection system can be improved as the amount of trace gas sample that can be present in a sample cell having the same optical path length increases. Thereby, a light absorption signal larger than the reference line can be generated. The pressure may be increased by increasing the flow rate while maintaining the volume of the sample cell.

  The field mirror surface and the objective mirror surface may be fixedly mounted so that the position of the field mirror surface and the objective mirror surface does not substantially change when the pressure is increased. For example, the field mirror surface and the objective surface may be mounted on a rod that holds these surfaces. Furthermore, the sample cell may be substantially airtight. The objective and field mirror surfaces in the sample cell may be exposed in the sample gas, thereby applying positive pressure to the back of each of the field mirror surface and objective surface to prevent deformation at high pressure. Can do. In various embodiments, this pressure may be between 1 atmosphere and about 10 atmospheres. In one embodiment, this pressure is 4 atmospheres.

  In some embodiments, signals at two different pressures may be measured and the ratio of these signals may be determined. This signal ratio allows for removal of baseline noise, increased sensitivity, and / or increased amplitude of the trace gas absorption profile relative to the baseline signal.

A first signal of a radiation beam propagating through a sample of ambient air at a first pressure in a sample cell is measured. The sample cell is pressurized with ambient air to a second pressure. A second signal of the radiation beam propagating through the ambient air sample is measured at the second pressure in the sample cell. A signal indicating the presence of a trace gas may be obtained by combining the first signal and the second signal. For example, these signals may be combined to generate an absorption profile for trace gases. In one embodiment, the radiation beam may include an interference signal. A trace gas absorption profile may be determined from the interference signal. In one embodiment, the first pressure is about 1 atmosphere and the second pressure is between about 1 atmosphere and 10 atmospheres. In one detailed embodiment, the first pressure is about 1 atmosphere and the second pressure is about 4 atmospheres.

  In various embodiments, the first signal is used as a baseline signal for the second signal because the optical alignment of the sample cell remains substantially unchanged when the pressure is increased. In some embodiments, the baseline signal is measured and used as a baseline signal for both the first signal and the second signal.

  In various embodiments, the flow system may include a cold finger for capturing the gaseous sample of interest by cooling the gaseous sample to a temperature below its saturation temperature. Many volatile substances condense below -75 ° C. In one embodiment, a cold cold trap is established at the gas outlet of the sample cell. After a specified period or collection period, one or several trapped gases may be rapidly vaporized by heating or “flashed back” into the sample cell, where spectral measurements are made. Also good. This technique can increase the amount of target gas by about one or two orders of magnitude while maintaining the sample cell at atmospheric pressure. In one embodiment, continuous flow measurements are taken at certain intervals, for example, approximately every 10 seconds, while flushing occurs at longer time intervals.

In various embodiments, the detection system may include a long wavelength pass filter. The noise due to the A / D converter may be comparable to the noise due to the detector. Incorporating a long wavelength pass filter can block higher wavenumber regions and improve sensitivity by reducing the digitizer dynamic range requirement by reducing the interferogram center burst value. Dynamic range of the detector without an optical filter may be between about 600 cm ^-1 to about 5000 cm ^-1. Since many of the toxic substances of interest can be detected with a value smaller than 1500 cm ⁻¹ , a spectrum higher than 1500 cm ⁻¹ can be eliminated by obtaining sensitivity using a long wavelength pass filter. For example, in a readily available standard long wavelength pass filter with a cutoff frequency of about 1667 cm ⁻¹ , the gain of the optical path length / NEA ratio may be about 20% to about 30%. In addition, using a long wavelength pass filter improves the signal-to-noise ratio of the detection system by allowing the detector to operate at a higher gain, for example, the highest gain achievable with a particular detector. Can do. In various embodiments, a low sensitivity detector such as an MCT detector or a DTGS detector may be used to record the spectrum in the higher frequency domain.

Biofuel may be used to drive the turbine generator engine. Biogas generally includes species such as, for example, hydrocarbons (eg, CH ₄ ) containing hundredths of the level of CO ₂ and H ₂ O. Biogas includes silane-containing hydrocarbons and siloxane compounds. Cyclic siloxanes (eg, D3-siloxane to D6-siloxane) may be present in the biogas produced by the digester. Biogas obtained from landfills may include linear siloxanes (eg, “linear” L2-siloxane to L6-siloxane), cyclic siloxanes, and / or trimethylsilanol (TMS). The concentration of TMS and siloxane compounds in the biogas may range from parts per million (ppm) levels to parts per billion (ppb) levels. TMS siloxane compound generates SiO ₂ particles together when oxidized in the turbine and promote excessive wear and cracking. Thus, the biofuel processing system may enable early detection and measurement of TMS and siloxane compounds by continuously monitoring the biofuel for TMS and siloxane. The system may use a stand-alone processor (eg, processor 34 of FIG. 1) to quantify the concentration of TMS and siloxane (eg, the level of siloxane impurities in the biofuel ranging from ppm to ppb levels). Stand-alone FTIR to detect).

  FIG. 9 shows a flowchart illustrating an exemplary method for monitoring silicon-containing compounds (eg, siloxanes) in biogas. The method includes supplying a non-absorbing gas (eg, nitrogen or helium) to a sample cell (eg, sample cell 22 of FIGS. 1 and 3) (step 205). Non-absorbing gases are gases that do not substantially absorb infrared radiation in the specified wavelength range of interest. The method also includes obtaining a first spectral measurement from the sample cell (eg, background instrument response) (step 210). Biogas is supplied to the sample cell (step 215). The biogas is at least 1 (eg, selected from the group consisting of TMS, L2-siloxane, L3-siloxane, L4-siloxane, L5-siloxane, D3-siloxane, D4-siloxane, D5-siloxane, or D6-siloxane). Contains two silicon-containing compounds. The method also includes obtaining a second spectral measurement from the sample cell (step 220). The first absorption spectrum based on the ratio of the first spectrum measurement value from the non-absorbing gas and the second spectrum measurement value (for example, the measurement value obtained from the sample gas composed of biogas supplied to the sample cell). Is generated (step 225). A second absorption spectrum is generated based on at least a first individual absorption spectrum for a known concentration of at least one silicon-containing compound in the biogas (step 230). The concentration of at least one silicon-containing compound (eg, siloxane or TMS) in the biogas is calculated by using the first absorption spectrum and the second absorption spectrum (step 235). For example, using CLS and / or other methods of performing a direct spectral comparison, at least one silicon after all possible interfering substances / gas are first removed from the spectrum (eg, the first absorption spectrum). The concentration of the contained compound can be calculated. Alternatively, the interfering substance / gas is not removed and is used in the spectral fitting routine.

  Both non-absorbing gas and biogas may be supplied to the sample cell (step 205 and step 215), thereby obtaining spectral measurements (step 210 and step 220). Biogas may be obtained, for example, from animal waste, waste water, or landfill. In general, the longer the data acquisition period (the period for obtaining a spectrum measurement), the lower the detection limit (eg, a lower concentration species can be detected). The longer the data acquisition period, the more accurate the measurement becomes possible (for example, the signal to noise ratio becomes higher). For example, if the noise is random (white noise), the signal to noise ratio increases with the square root of the acquisition period. The second spectral measurement (eg, step 220) may be obtained over an acquisition period of about 10 seconds to about 90 seconds. In some embodiments, the second spectral measurement from the sample cell is obtained in a wavelength range from about 8 microns to about 12 microns. Obtaining the second spectral measurement may include obtaining an infrared signal from the sample cell (eg, obtaining a sample of a gas composed of biogas).

The concentration of the at least one silicon-containing compound is determined in real time (eg, as a result of quantifying the concentration of TMS or siloxane compound obtained in seconds or minutes) on-site (eg, in-line or biogas source and fluid It may be calculated (eg, step 235) in a communicating device and without the need for a container or absorbent medium to collect the sample gas. The sample cell and processor (eg, processor 34 of FIG. 1) may be in fluid communication with a source of biogas, and the sample is obtained (eg, by existing GC / MS methods) and transported from the field for analysis. The analysis may be performed at / substantially near the source without having to do so. The time for obtaining and analyzing the sample (eg calculating the concentration of siloxane, the concentration of TMS, and / or the concentration of all silicon-containing compounds in the sample) allows the silicon-containing compounds to be accurately measured at the level in the biogas mixture. It may be on the scale of seconds to minutes depending on the final signal-to-noise ratio required to quantify. For example, if the signal to noise ratio is not sufficient to accurately measure a particular concentration, the acquisition time can be extended to further reduce noise (eg, increase the signal to noise ratio). In some embodiments, the turbine generator is shut down when the concentration of the at least one silicon-containing compound reaches a threshold value.

  In some embodiments, a processor (eg, processor 34 of FIG. 1) may be used to calculate the concentration of at least one silicon-containing compound in the biogas (step 235). Chemometrics combined with spectroscopy (eg, FTIR spectroscopy) and mathematics (eg, multiple regression analysis) may generate clear quantitative information about silicon-containing compounds in biogas. For example, using a processor, a multiple regression analysis is performed using the first absorption spectrum and the second absorption spectrum to calculate the concentration of at least one silicon-containing compound in the biogas. Multiple regression analysis may be performed using classical least squares (CLS), partial least squares (PLS), inverse least squares (ILS), principal component analysis (PCA), and / or other chemometric algorithms. Good.

  The second absorption spectrum (eg, obtained in step 230) includes at least a first individual absorption spectrum and one or more additional siloxane compounds (eg, L2-siloxane, L3-siloxane, L4-siloxane). , L5-siloxane, D3-siloxane, D4-siloxane, D5-siloxane, or D6-siloxane), silicon-containing alcohols such as trimethylsilanol (TMS), for example, hydrocarbon compounds including aromatic compounds and chlorinated hydrocarbons, It may be generated based on individual absorption spectra for water or carbon dioxide. The second absorption spectrum may be a model based on a known concentration of a siloxane compound, TMS, hydrocarbon compound, water, or carbon dioxide (eg, a model representing an individual absorption spectrum of a substance in biogas). . In some embodiments, the second absorption spectrum is at least a first individual absorption spectrum (eg, for a siloxane compound) and / or one or more additional siloxane compounds, TMS, hydrocarbon compounds (eg, , Methane or ethane), water, or carbon dioxide.

  In some embodiments, the concentration of at least one siloxane compound such that the second absorption spectrum is substantially similar to the first absorption spectrum (eg, mathematically fitting the model absorption spectrum to the measured absorption spectrum). Is determined (eg, step 235). As an example, at least one variable representing the concentration of at least one siloxane compound is provided such that the second absorption spectrum is substantially similar to the first absorption spectrum (eg, the second absorption spectrum is the first The concentration of at least one siloxane compound may be calculated by determining the value of at least one variable (eg, the concentration value) that is mathematically applied to the absorption spectrum of

  For example, a variety of different types of quantitative analysis based on both univariate and multivariate analysis techniques may be used to link spectral measurements directly to the actual chemical composition. The univariate method involves correlating the spectral peak height or region below the spectral curve with the same characteristics for the known stoichiometry of the species in the biogas. In some embodiments, this may be done, for example, by developing a quantitative model that predicts the actual concentrations of various species in the biogas using least squares regression. Another univariate method that can be used in alternative embodiments is a K-matrix or classical least squares (CLS) based on an explicit linear additive model (eg, Beer's Law described in Equation 1 below). CLS uses a larger portion of the spectrum (or the entire spectrum) in regression for all chemical components in the spectral region.

CLS reduces model accuracy, for example, due to unknown concentrations, so the concentrations of all components that are spectrally active are known and included in the calibration model so that an appropriate predictive model can be developed. There is a restriction that it is necessary. To avoid this and other problems that may arise when using univariate models, multivariate techniques are usually more useful. One multivariate method uses multiple linear regression (MLR) (also called P matrix or inverse least squares (ILS)) to create a model using only the concentration of the chemical component of interest (eg, H. Mark). (See H. Mark) Analytical Chemistry, 58, 2814, 1986). With this technique, a model can be created without unwanted effects using only known concentrations, but this model limits the number of wavelengths that can be used to describe each component.

  (As described in Fredericks et al. Applied Spectroscopy, 39: 303, 1985) (as well as the ability of the CLS model) to represent compositions of interest using large regions of the spectrum. Other multivariate techniques that combine the ability to address only the composition of interest (as well as the capabilities of the MLR model) may be used in alternative embodiments. In one embodiment, principal component regression (PCR) is used (as described in Fredericks et al. Applied Spectroscopy, 39: 303, 1985). This method is based on performing spectral decomposition using principal component analysis (PCA) and then performing a regression of known concentration values against the PCA score matrix. Specifically, when PCR is used, the PCA is first configured by an X matrix that yields a score matrix T and a loading matrix P. The next step uses some of the first score vectors in the multiple linear regression with Y data. If the first few components of PCA are components that actually summarize most of the X information for Y, PCR works as well as partial least squares (PLS) for stereo data. This will be described below.

  In another embodiment, PLS can be used to obtain the actual concentration value of the lesion composition based on the spectral data (eg, W. Lindberg, J. W. Lindberg, J. Parson, and S. Wald). Persson, and S. Wald) Analytical Chemistry, 55: 643, 1983; P Geladi and B Kowalski, Analytical Chemical Acta, 35: 1, 1986; and Harland, Thomas Chemistry, 60: 1193, 1202, 1988). PLS is similar to PCR, but when using PLS, both spectral and concentration information are decomposed at the start of the method, and the resulting score matrix is exchanged between the two groups. This gives heavier weight in the model to the spectral information correlated with the concentration information, thus providing a more accurate model than PCR. The core of the PLS algorithm is non-linear iterative partial least squares (NIPALS) (eg, Wald Perspectives in Probability and Statistics, J Gani) (Academic Press, London, pp 520-540, pp. 540-540). ) Or simple partial least squares (SIMPLS) (Jong Chemon. Intel. Lab. Syst., 18: 251, 1993) algorithm.

For further details of PCA analysis, PCR analysis, MLR analysis, and PLS analysis, see “Multi- and Megavariate Data Analysis, Part I, Basic Principles and Applications” Ericsson et al., Ericsson U.S. , January 2006 and "Multi- and Megavariate Dat"
a Analysis, Part II, Advanced Applications
and Methods Extensions ", Eriksson et al., Umetrics Academy, March 2006.

  As described above, various chemometric algorithms (eg, PCA, PCR, MLR, PLS) may be used to calculate the concentration of one or more silicon-containing compounds in the biogas. The chemometric algorithm method fits the overall absorption rate (eg, a measured spectrum based on spectral measurements obtained from biogas) to the absorption rate of each composition type (eg, siloxane, TMS, methane) Used to determine the calculated concentration of the seed. Beer's law is as follows:

Is wave number

The absorbance of species i at

Is the absorption rate of the species in the wave number, b is the optical path length, the concentration of c _i species. Thus, by measuring the absorbance of a species at a known concentration, it is possible to determine the absorbance of the species for a known concentration and a given wavelength (eg, wave number). An absorption spectrum can be generated by measuring the absorbance of a species at a known concentration for a range of wavelengths.

  If there are multiple species (eg, molecules) in the sample, Equation 1 reflects that the measured absorbance of the sample (eg, sample gas in the sample cell) is the sum of the absorbances of all species in the sample. It may be modified as follows. As an example, if the biogas contains one or more silicon-containing compounds, hydrocarbon compounds including aromatics and chlorinated hydrocarbons, water, and carbon dioxide, the measured absorbance of the biogas sample is the species in the biogas. The sum of all absorbances of (eg, silicon-containing compounds, hydrocarbon compounds, water, and carbon dioxide). Thus, quantitative analysis may be used to predict the actual concentration of various silicon-containing compounds in the biogas.

A chemometric algorithm may be used to determine the concentration of species in the sample. For example, using the chemometric algorithm in Equation 1 and / or other equations, the model spectrum (eg, by mathematically fitting the model spectrum to the measured spectrum after all interference components have been removed) A concentration value such that the second absorption spectrum is substantially similar to the measurement spectrum (for example, the first absorption spectrum) may be obtained.

  In one embodiment, PLS is used to calculate the concentration of siloxane (and / or other compounds in the biogas). FIG. 10 is a diagram representing a non-linear iterative partial least squares (NIPALS) decomposition of spectral information represented by a matrix X containing spectral measurements and a matrix Y containing concentration information. The PLS component of the PLS model is calculated conventionally using the NIPALS algorithm (or other similar decomposition algorithm). In PLS, two data matrices X and Y are related to each other by a linear multivariate model. In short, the linear model specifies the relationship between a dependent variable or response variable y or a set of response variables Y and a set of predictor variables X. For example, the response variable y is the concentration and the predictor variable X is the spectrum measurements 1002a to 1002n. The numbers 1.0 and 0.45 in Y are the calculated concentrations of gas components in the corresponding spectrum. There are many variations of the NIPALS algorithm consisting of matrix-vector multiplication (eg, X'y). S and U are score matrices obtained from spectrum information and component information, respectively. The numbers 0.39 and -0.37 in S are basic vector scalar (score) modifiers representing the linear combination of the first set of spectra. These numbers are only examples of the number of first rows that satisfy S and U. In this example, the entire set of observed spectra is decomposed into two basis vectors, so there are two numbers. When these numbers are multiplied by the corresponding lines in the PCx representation, the first spectrum is regenerated (eg, minimum noise). In other embodiments, a set of observed spectra may be decomposed into any number of basis vectors. PCx and PCy are principal components (or latent variables / eigenvectors) obtained for spectral information and component information, respectively. PCx includes latent variables 1004a to 1004f. Other symbols in the figure are the number of spectra (n), the number of data points per spectrum (p), the number of components (m), and the number of last latent variables / eigenvectors (f).

  Generate a latent variable and score for each of the X and Y matrices by a first decomposition of spectral data and concentration / composition data, and replace the score matrix of spectral information (S) with a score matrix containing concentration information (U) . Subtract the latent variables obtained from PCx and PCy from the X and Y matrices, respectively. These newly subtracted matrices are then used to calculate the next latent variable and score for each round until enough latent variables are found from PCx and PCy to represent the data. Before each decomposition round, the new score matrix is exchanged and the new latent variables obtained from PCx and PCy are removed from the reduced X and Y matrices.

  Since the spectrum information is correlated with the concentration information when the score matrix is exchanged, the final number (f) of latent variables (or basis vectors) obtained from the PLS decomposition is highly dependent on the concentration information by the exchanged score matrix. Correlated. Advantageously, exchanging the score matrix leaves basis vectors in both sets of matrices that are naturally correlated with each other. The PCx and PCy matrices include spectral variations that are highly correlated to the composition used to create the model. The second set of matrices, S and U, contain the actual scores that represent the amount of each of the latent variable variants present in each spectrum. In the PLS model, S matrix values are used.

  In one embodiment, the PLS method is used to predict the actual components of the siloxane compound in the biogas. For example, the PLS algorithm can be used to directly predict the chemical content of biogas or, for example, existing compounds (eg, hydrocarbon compounds including silicon-containing compounds, aromatic compounds and chlorinated hydrocarbons, water, (Or carbon dioxide) as a percentage.

In another embodiment, CLS is used to calculate the concentration of siloxane (and / or other compounds) based on the model spectrum and the measured spectrum. In some embodiments,
The sample contains two components and / or species (s ₁ and s ₂ ) in the mixture. The biogas may include more than two components / species (eg, the biogas may include various types of silicon-containing compounds, hydrocarbon compounds, etc.). However, for clarity of explanation, the following example assumes two components.

If the sample contains two species, the species should have at least two different wave numbers. In one embodiment, CLS can be used to model the absorbance of two wave numbers based on the relationship of absorbance at each wavelength. For example, the absorbance of the first wavelength is the absorbance of the first species s ₁ at the first wavelength, the absorbance of the second species s ₂ at the first wavelength, the optical path length (eg, FIG. the optical path length of the sample cell 22 as described with respect to 4), a first concentration of species s _1, based on the relation of the residual obtained by regression analysis of the second concentration of the species s _2, and the first wavelength. Similarly, for example, the absorbance of the second wavelength is the absorbance of the first species s ₁ at the second wavelength, the absorbance of the second species s _{2 at} the second wavelength, the optical path length, the first species s. _{Based on} the concentration of the _first and second species s ₂ and the residual relationship obtained by regression analysis of the second wavelength.

When the optical path length is constant, the optical path length is not considered when determining the absorbance of each wavelength. Instead, the absorbance at the first wavelength is the absorption coefficient of the first species s ₁ at the first wavelength, the absorption coefficient of the second species s ₂ at the first wavelength, and the first species s ₁ . Based on the relationship between the concentration of the second species s ₂ and the residual obtained by regression analysis of the first wavelength. Similarly, the absorbance of the second wavelength is the absorption coefficient of the first species s ₁ at the second wavelength, the absorption coefficient of the second species s _{2 at} the second wavelength, and the first species s ₁ and a second species s ₂ concentration, based on the relationship between the residual obtained by regression analysis of the second wavelength.

Using the above relationship, the absorption coefficient can be determined for a certain wavelength by measuring the absorbance of the sample at a known concentration. Such absorption coefficients can then be used to measure / determine the unknown concentration of species s ₁ and s ₂ in the sample. For example, the absorbance (for example, measurement spectrum) of the sample may be measured at two wavelengths, and the absorbance value of each wave number may be obtained. Since the extinction coefficient is known, it may be used together with the absorbance value to calculate the concentration of the species.

  As mentioned above, the biogas may contain more than two components / species. In such cases, absorbance, absorption coefficient, and concentration values may be modeled using the following matrix.

  In the above equation, “A matrix” is a matrix of spectral absorbance, “K matrix” is a matrix representing an extinction coefficient, and “C matrix” is a matrix representing a concentration. The number of samples (spectra) is represented by “n”, the number of wavelengths used for calibration is represented by “p”, and the number of species / components is represented by “m”. Equation 6 may be simplified and used to calculate the concentration of the species in the sample.

C = A · K− ¹ Formula 3
In the above equation, K ⁻¹ is the inverse of the K matrix. The K matrix of Equation 2 may be solved by measuring the absorbance of a sample with known concentration of each species and using the following equation:

K = A · C ⁻¹ Formula 4
If the concentration of an individual species (eg, siloxane compound, hydrocarbon compound, water, or carbon dioxide present in biogas, for example) is known, the “C matrix” is known. The “A matrix” may be created based on spectral measurements obtained using, for example, the detection system of FIG. 1 (eg, FTIR spectrometer). Therefore, if the A matrix and the reciprocal of the C matrix with a known density are used, the K matrix is obtained using Equation 4.

  After calculating the K matrix from Equation 4, the concentration in the sample is calculated using Equation 3. Using the inverse of the K matrix (e.g., calculated from Equation 4 using a sample with a known concentration), the species in the sample where the concentration of the individual species is unknown (e.g., siloxane in biogas ) Is calculated. A detection system (eg, the system of FIG. 1) may be used to obtain a spectral measurement of a sample (eg, sample biogas). An A matrix that represents a summary of the absorbance of individual species in the sample is generated based on the spectral measurements. The inverse of the K matrix and the A matrix are used in Equation 3 to calculate the concentration of individual species in the sample.

  FIG. 11 shows a CLS of a wastewater digestion gas compound containing 920 ppb D4-siloxane, 400 ppb D5-siloxane, 65% methane, 35% carbon dioxide, 1400 ppm ethane, 340 ppm propane, and 65 ppm butane (ie, It is a graph of an analysis result of classical least squares. The graph shows absorbance values (y-axis) as a function of wavelength (ie, wavenumber) (x-axis). Curve 300 represents the measured spectrum, curve 305 is the individual absorption spectrum of methane, curve 310 is the individual absorption spectrum of carbon dioxide, curve 315 is the individual absorption spectrum of ethane, and curve 319 is the propane spectrum. Individual absorption spectra, curve 320 is the individual absorption spectrum of butane, curve 325 is the individual absorption spectrum of D4-siloxane, and curve 330 is the individual absorption spectrum of D5-siloxane. Since the model absorption spectrum overlaps with the measurement spectrum 300 because the residual value is relatively small, the model absorption spectrum representing the sum of the composition spectra 305, 310, 315, 320, 325, and 330 is used to clarify the figure. Not shown.

  You may calculate the density | concentration of a siloxane compound (for example, D4-siloxane and D5-siloxane) using data, such as a spectrum shown in FIG. Using the measured / observed spectrum 300, the values of the A matrix of Equation 7 and Equation 10 may be added. Individual absorption spectra 305, 310, 315, 320, 325, and 330 for individual species for known concentrations may be used to add values for K and / or P matrices of Equation 7 and Equation 10. . Therefore, the measured A matrix and the calculated K matrix and / or P matrix are used to determine the known concentration values of individual species in the measured spectrum.

In another embodiment, ILS may be used to calculate the concentration of the species in the sample. In CLS, absorbance is a dependent variable. In ILS, concentration becomes a dependent variable. For example, the concentration of the first species s ₁ is the linear inverse coefficient (which is a function of the absorption rate of the first species s1 at two wavelengths), the absorbance at the first wavelength and the second wavelength, Based on the relationship with the residual obtained from the regression analysis of _one species s1. This is simplified to the following matrix when there are several species in the sample.

C = P · A + E _C Formula 5
In Equation 5, C is a concentration matrix, P is a linear inverse coefficient, A is an absorbance matrix, and E is a residual matrix. As with CLS, the P matrix may be determined using a known concentration of the sample. In this scenario, the ILS model can be recalculated until the residual is close enough to zero (eg, by setting a threshold indicating that the residual is close enough to zero), and Since Equation 5 can be modified as follows, the residual may be assumed to be zero.

P = C · A ⁻¹ Formula 6
Using the known concentrations of the individual species, the value of the C matrix is obtained. The A matrix is created based on spectral measurements obtained from a sample having a known concentration (eg, using the detection system of FIG. 1 such as an FTIR spectrometer). Thus, the P matrix may be calculated using Equation 6 based on a spectrum measured from known concentrations of species in the sample.

  The P matrix may then be used in Equation 5 to solve for the unknown concentration of species in the sample. In particular, a detection system (eg, of FIG. 1) such as an FTIR system may be used to obtain a spectral measurement from a sample having an unknown concentration of an individual species. Spectral measurements may be used to add absorbance values in the A matrix. The P matrix is calculated using Equation 6 based on the known concentrations, and the P matrix is used in Equation 5 to calculate the C matrix, thereby obtaining concentration values for individual species in the sample. Good.

  A system, such as the system of FIG. 1 above, incorporating any of the exemplary techniques as described above may be used to quantify and monitor silicon-containing compounds (eg, siloxanes) in biogas. The system includes, for example, a light source of a first radiation beam (eg, light source 14 of FIG. 1), an interferometer (eg, interferometer 18 of FIG. 1), and a sample cell (eg, cell 22 of FIG. 1). A flow mechanism (eg, flow system 82 of FIG. 3), a cooling detector (eg, detector 30 of FIG. 1), a processor (eg, processor 34 of FIG. 1), a light source, an interferometer, a sample cell, It may include a cooling detector and a housing (eg, housing 42 of FIG. 1) in which the processor is disposed. The interferometer receives a first beam of radiation from a light source and is reflected a second beam of radiation (eg, an interferometer signal) (eg, about 48 times total in the sample cell, about 10.18 meters). A second beam that produces an effective optical path length of The sample cell is in optical communication with the interferometer. The flow mechanism is a flow of non-absorbing gas (eg, a gas that does not substantially absorb infrared radiation in the specified wavelength range of interest) and a biogas (eg, pressurized gas introduced into the sample cell). A second flow of sample (eg, 3 psig to 5 psig) (eg, 400 mL biogas), biogas convection time is on the order of about 5 seconds). A detector (eg, a cooled detector) is in optical communication with the sample cell, and a first interference signal that propagates in the non-absorbing gas in the sample cell and a second that propagates in the sample gas in the sample cell. The sample gas is made of biogas. The processor is electrically connected to a detector (eg, a low temperature (eg, Stirling ™ engine) cooled MCT (cadmium mercury telluride) detector) to calculate the concentration of at least one siloxane compound in the biogas. . The processor calculates a concentration of at least one siloxane compound in the biogas based on the first absorption spectrum and the second absorption spectrum using chemometric techniques (eg, CLS technique and ILS technique). The first absorption spectrum is based on the ratio of the first interference signal and the second interference signal obtained from the detector. The second absorption spectrum is based at least on an individual absorption spectrum for a known concentration of at least one siloxane compound.

  In some embodiments, the sample cell (eg, cell 22 of FIG. 1 having the optical configuration described above with respect to FIG. 2) is a concave reflective field mirror surface (eg, FIG. 1) at the first end of the sample cell. Two field mirror surfaces 78) and a substantially spherical concave reflective objective mirror surface (eg, objective mirror surface 74 in FIG. 2) at the second end of the sample cell opposite the field mirror surface. And the objective surface includes a focus in at least one plane to maximize the throughput of the second radiation beam propagating through the sample cell via a plurality of reflections at each of the field mirror surface and the objective surface. It has a cylindrical component that increases the degree of coincidence.

  In one embodiment, the computer readable product is tangibly embodied on an information carrier or machine readable storage device and is a digital signal processor (eg, processor 34 of FIG. 1) of a biogas detection system (eg, the system of FIG. 1). It can be operated on. The computer readable product receives a first spectral measurement (eg, from detector 30 of FIG. 1) obtained from a non-absorbing gas in a sampling cell (eg, cell 22 of FIG. 1) in a digital signal processor. The non-absorbing gas substantially does not absorb infrared radiation in the specified wavelength range of interest. The computer product causes a digital signal processor to receive a second spectral measurement obtained from a sample gas consisting of biogas in the sampling cell and based on a ratio of the first spectral measurement to the second spectral measurement. A first absorption spectrum (for example, a measured absorption spectrum) may be generated. A second absorption spectrum (eg, a model absorption spectrum) may be generated / formed based on at least a first individual absorption spectrum for a known concentration of at least one siloxane compound. The computer product may cause the processor to calculate the concentration of the one or more siloxane compounds using the chemometric techniques described above (eg, performing a multiple regression analysis and extracting the second absorption spectrum to the first absorption spectrum). Mathematically fit to the spectrum to calculate the concentration of at least one siloxane compound in the biogas).

  As noted above, absorption spectra based on spectral measurements of the biogas in the sample cell and individual spectra for individual components / species in the biogas (eg, siloxane and silicon-containing compounds, hydrocarbon compounds, water, or The individual spectra of species in biogas such as carbon dioxide) may be used to calculate the concentration of individual species in the sample. A collection of individual absorption spectra (eg, based on concentration ranges and / or various spectral mixtures depending on the analytical method or methods used) using an absorption spectrum such as that shown in FIG. A model may be generated that is based on a calibrated absorption spectrum that represents (eg, the second absorption spectrum described above with respect to FIG. 9). Specifically, an A matrix as described above is generated using an absorption spectrum based on a spectrum measurement value of an unknown biogas, and the concentration of siloxane can be calculated using, for example, Equation 7 and Equation 11. Good. Individual spectra obtained based on measurements with known concentrations of species may be used to generate model spectra representing individual species. Individual spectra for known concentrations may be used to calculate P or K matrices as described above in Equations 8 and 12. This model may include, for example, a P matrix or K matrix (eg, determined by using known concentrations of species) that can be used to calculate the concentration of siloxane using, for example, Equation 7 and Equation 11. Good. FIG. 11 shows a spectrum of data that can be used to quantify the concentration of siloxane compounds in biogas, according to an illustrative embodiment of the invention. Each absorbing species in the sample has a unique absorption versus frequency distribution (ie, an absorption spectrum). Using chemometric algorithms (eg, multiple regression analysis), each component can be characterized and quantified, thus even in the presence of interfering absorbers (eg, hydrocarbon compounds such as methane or ethane). Individual species of siloxane compounds can be detected.

  In general, the method for monitoring silicon-containing compounds in biogas described above with respect to FIG. 9 is based on a second absorption spectrum based on the individual absorption spectra of all known silicon-containing compounds and hydrocarbon compounds having known concentrations. Generating an absorption spectrum of (step 230). Thus, the second absorption spectrum represents all possible cyclic and linear siloxane compounds and other silicon-containing compounds such as TMS that can be present in a given sample (eg, speciation techniques). . Using this seeding technique, for example, the CLS (Classical Least Squares) analysis described above with respect to Equations 2 to 4 or the ILS (ie, inverse least squares) analysis described above with respect to Equations 5 and 6 is used. Thus, the concentration of various types of silicon-containing compounds (eg, siloxane and / or TMS) in the sample may be calculated (step 235).

  One drawback of using speciation techniques is that low level concentrations (eg, less than 0.02 ppm-v) of silicon-containing compounds in the sample cannot be detected. For example, if all known siloxane compounds are used in the first absorption spectrum when modeling, but at least one siloxane is not present in the actual sample or is similar to other components, In particular, accurate measurement of various species of siloxanes in samples at low concentrations can be hindered. Another disadvantage of the speciation technique is that all unknown silicon-containing compounds that can be present at any time during the analysis need to be included as part of the second absorption spectrum. The resulting cross-correlation can inject noise throughout the various analyzes that contribute to the concentration of siloxanes, TMS, and other silicon-containing compounds, thereby reducing the ability to perform chemical detection at low ppb levels. There is sex.

  In view of these shortcomings, another method for detecting / monitoring silicon-containing compounds in biogas samples (landfill gas or digestion gas) is provided. In this method, instead of calculating the concentration of each silicon-containing compound present in the sample, one or more total concentration values are calculated. The total concentration value may be, for example, a single value for the total concentration of siloxanes in the sample, a single value for the total concentration of other silicon-containing species in the sample, and / or a single value for the total concentration of all silicon-containing species. One value may be included. A single value, unlike the speciation method, corresponds to a subset of these known compounds rather than using all of the silicon-containing and / or hydrocarbon compounds that are commonly present in the biogas of interest. May be determined based on one or more absorption spectra. Specifically, the method uses known concentrations of all silicon-containing compounds and / or hydrocarbon compounds to fit a biogas sample based on, for example, CLS as used in speciation methods. Rather, the fitting analysis is performed using only a selected subset of known silicon-containing compounds and / or hydrocarbon compounds.

  FIG. 12 shows a flowchart 300 illustrating another exemplary method for monitoring silicon-containing compounds in a biogas sample. The method includes supplying a non-absorbing gas (eg, nitrogen or helium) to a sample cell (eg, sample cell 22 of FIGS. 1 and 3) (step 305). A non-light-absorbing gas is a gas that does not substantially absorb infrared radiation in the specified wavelength range of interest. The method also includes retrieving a first spectral measurement from the sample cell (eg, background instrument response) (step 310). Biogas is supplied to the sample cell (step 315). The biogas has an unknown concentration (eg, obtained from the group consisting of L2-siloxane, L3-siloxane, L4-siloxane, L5-siloxane, D3-siloxane, D4-siloxane, D5-siloxane, or D6-siloxane). ) At least one silicon-containing compound such as at least one siloxane compound; The method also includes obtaining a second spectral measurement from the sample cell (step 320). The first absorption spectrum includes a first spectrum measurement value obtained from a non-light-absorbing gas and a second spectrum measurement value (for example, a measurement value obtained from a sample gas composed of biogas supplied to a sample cell). Generated based on the ratio (step 325). In some embodiments, steps 305, 310, 315, 320, and 325 of flowchart 300 of FIG. 12 are substantially the same as steps 205, 210, 215, 220, and 225 of flowchart 200 of FIG. 9, respectively.

Still referring to FIG. 12, a surrogate set of absorption spectra is determined (step 330). Each absorption spectrum in the surrogate assembly is based on the individual absorption spectra of known silicon-containing compounds, but unlike the speciation method described above with respect to FIG. 9, all known silicon-containing compounds are included in this surrogate. It is not always used in the method. Here, a surrogate method for generating a surrogate set of absorption spectra is used. The surrogate approach involves selecting only a subset of known silicon-containing compounds (eg, siloxane compounds) to be used in the modeling phase. This subset is, for example, D
It may be selected from a larger collection including 3-siloxane, D4-siloxane, D5-siloxane, D6-siloxane, L2-siloxane, L3-siloxane, L4-siloxane, L5-siloxane, and trimethylsilanol (TMS). The absorption spectrum for a subset of silicon-containing compounds selected from a larger group of known silicon-containing compounds is hereinafter referred to as a surrogate set. In some embodiments, only 3 to 5 siloxane compounds from all known silicon-containing compounds are selected as siloxane compounds to be included in the surrogate assembly. Only one siloxane compound may be included in the surrogate assembly. In some embodiments, the surrogate set consists of a subset of a larger set of known hydrocarbon compounds including, for example, methane, ethane, butane, propane. In another example, a larger collection of known hydrocarbons may include methane, toluene, ethanol, and methanol. In some embodiments, the surrogate set consists of a subset of known siloxane compounds and / or hydrocarbon compounds. In some embodiments, the surrogate set consists of a subset of known siloxane compounds, known hydrocarbon compounds, and / or TMS. In general, the surrogate set of absorption spectra used in the model is based on the individual absorption spectra of each compound in a subset of known silicon-containing compounds (including siloxanes and TMS) and / or hydrocarbon compounds.

  Which silicon-containing compound and / or hydrocarbon compound is selected as the silicon-containing compound and / or hydrocarbon compound to be included in the surrogate assembly may depend on the type of sample. For example, in the case of landfill biogas, surrogate aggregates are: a) L2-siloxane, L3-siloxane, and D4-siloxane, b) L2-siloxane, D3-siloxane, and D4-siloxane, or c) L2-siloxane, D3-siloxane and D5-siloxane may also be included. In some embodiments, TMS is also added to the surrogate set depending on the landfill years. In the case of digestive biogas, surrogate assemblies include a) D3-siloxane, D5-siloxane, and L3-siloxane, b) D4-siloxane, D5-siloxane, and L3-siloxane, or c) D3-siloxane, D5- Siloxane and L2-siloxane may also be included.

Similar to the siloxane compound, one or more hydrocarbon compounds in the surrogate assembly may be selected based on the type of biogas sample. For example, the hydrocarbon compound selected can depend on which hydrocarbon compound is likely to be present in the biogas. In the case of landfill biogas, the hydrocarbon compounds in the surrogate assembly include, but are not limited to, about 10 ppm up to about 1% level (and possibly higher levels) of ethanol, methanol, Toluene and / or freon ™, up to 95% methane, and / or from about 5% up to about 50% CO ₂ may be included. In the case of digestion gas, the hydrocarbon compounds in the surrogate assembly include, but are not limited to, ethane, propane, and / or about 100 ppm up to about 1% level (and possibly higher levels) of hydrocarbon compounds. Alternatively, butane and / or about 10% up to about 50% CO ₂ may be included.

Which silicon-containing compound and / or hydrocarbon compound is selected as the silicon-containing compound and / or hydrocarbon compound to be included in the surrogate assembly of a specific sample is determined by experiment, for example, the spectral peak of the selected surrogate compound is a biogas. It may be determined based on whether it matches one or more of the spectral peaks of the compound in the sample. For example, the surrogate set of compounds may be determined by examining gas components using GC / MS. If the biogas sample is already pressurized, the pump is not necessary. In one embodiment, the FTIR spectrometer operates at a scan rate of 10 to 20 seconds and an average scan count of 36 to 72 times. This can make the detection limit very low. The biogas sample may be input to the gas cell of the FTIR spectrometer without being heated. The gas cell may be heated at 40 ° C., for example. The spectrum may be measured from 35 ° C to 40 ° C. The spectrum in the sample may be collected with a resolution of about 4 cm ⁻¹ using a gas cell path length of about 10.18 m to lower the detection limit. Methane may be flowed through the FTIR spectrometer and the observed methane spectrum in biogas may be used to calibrate each system ranging from 40% up to 100%. Thereby, a good spectral subtraction result can be obtained. For example, even in the presence of high methane levels, the detection limit of siloxane concentration and / or total silicon concentration can be lowered to estimate very little residual and small spectral details.

  In some embodiments, the silicon-containing compound or hydrocarbon compound selected as the silicon-containing compound or hydrocarbon compound to be included in the surrogate assembly is likely to be present in the biogas sample, but this is not required. That is, the silicon-containing compound or hydrocarbon compound in the surrogate assembly need not be present in the sample. In some embodiments, a library of surrogate sets may be maintained, and each surrogate set may be used for a particular type of biogas sample. For example, a surrogate set comprising L2-siloxane, D3-siloxane, and D4-siloxane may be used to model any landfill biogas regardless of when / where the biogas is collected.

  The first absorption spectrum is determined (step 325), the surrogate set of the absorption spectrum is determined (step 330), and the total concentration of all siloxane compounds in the biogas by using the first absorption spectrum and the surrogate set of the absorption spectrum Is calculated (step 335). The total concentration of all silicon-containing compounds may be calculated. Using CLS, PLS, ILS, PCA, and / or other methods that perform direct spectral comparisons, after interfering substances / gases have been removed via modeling methods (eg, methane, CO2, specific A single number representing the total concentration of all siloxane species and / or silicon-containing species in the sample (which removes interfering hydrocarbons, etc.) can be determined. The total concentration may be determined such that the surrogate set of the absorption spectrum is compared with the first absorption spectrum, and the spectral feature difference determined from the fitting routine is minimized. In some embodiments, at least one of a baseline value or an offset value is also used as a fitting parameter in the fitting routine. As an example, in the CLS analysis described above with reference to Equations 2 to 4, in consideration of the fact that the number of compounds used in the surrogate set is smaller, the surrogate method uses A as compared to the speciation method. Reduce the size of the matrix, K matrix, and C matrix. In some embodiments, all of the resulting respective concentration values for each surrogate siloxane in the C matrix are summed to calculate the total siloxane value to calculate the total concentration of siloxane in the biogas sample. This total siloxane value may be expressed as ppm or mg / m 3. In some embodiments, a total silicon value may be calculated that represents the total concentration of silicon-containing compounds in the biogas sample. This value may be determined by first correcting each surrogate concentration by the amount of silicon present in that particular component when the surrogate contains silicon molecules. The TMS surrogate, other silicon-containing surrogate, and siloxane surrogate concentrations in the C matrix are then summed to produce the total silicon value. In some embodiments, the concentrations may be summed individually as a silicon-containing component to produce a total silicon value and a total siloxane value as a siloxane-containing component. In general, one of ordinary skill in the art will use a surrogate approach based on the analytical technique (eg, CLS, PLS, ILS, or PCA) described above with reference to the speciation technique and / or the total concentration of siloxane compounds in the sample and / or It can be easily determined how to calculate the total concentration of the silicon-containing compound.

Method 300 may include an optional step (step 340) of applying a correction factor to the total siloxane value and / or total silicon value determined in step 335, such as scaling the value by a particular factor. The correction factor may be determined based on the type of biogas being analyzed, the siloxane compound used in the surrogate assembly, the hydrocarbon compound used in the surrogate assembly, or any combination thereof. For example, the correction factor may be calculated by comparing a biogas containing a known concentration of a siloxane compound with the siloxane concentration determined in step 335 for the same biogas. Based on the comparison, it may be determined whether a correction factor is necessary, and if so, it may be determined what the correction factor should be. In some embodiments, after correction factors are determined for certain biogases such as landfill gases, the same correction factors may be used for analysis of other landfill gases. In some embodiments, different correction factors are used for different surrogate sets. In some embodiments, the correction factor is modified for each biogas sample, for example, as a result of a field test.

  Dilution tests have shown that the surrogate technique works better than the speciation technique. However, there are a number of advantages associated with the surrogate approach. One advantage is that the surrogate approach allows on-site process monitoring at various landfills without the need to change or modify the method. Another advantage is that the sample is not processed (eg, by capturing the sample in a portable container or impinging the gas stream into the alcohol solution), and thus compared to when the sample is processed. Thus, the siloxane compound in the sample can be quantified more accurately. Another advantage of the surrogate approach is that no canister and / or sample bag is required to obtain a sample from the biogas stream. For example, it may be connected directly to the biogas stream through a sample pipe that does not absorb siloxane. The biogas may be transferred directly to the FTIR analyzer via the sample piping.

  In addition, the surrogate technique reduces the total siloxane value from 600 ppb to 60 ppb, such as reducing the overall detection limit of siloxane compounds and / or silicon-containing compounds in biogas compared to the speciation technique of FIG. Can only be lowered. In some embodiments, for example, FTIR analysis can be used to achieve a single digit detection limit. It will be apparent to those skilled in the art that FTIR analysis includes Fellgett and Jacquinot gains (eg, spectral throughput and sensitivity are naturally improved). Combined with the gain of such Fellgett and Jacquinot and the use of a high-sensitivity detector cooled at low temperature, high-quality detection can be performed.

  In some embodiments, a surrogate approach may be used to monitor siloxane levels and / or total silicon levels present in landfill gas or digestion gas produced after the filtration system. The gas stream may be used to drive turbines, boilers, automobiles, and / or home appliances, all of which have not been monitored and adjusted for siloxane content and / or silicon content There is a risk of damage. Siloxane levels and / or silicon levels may need to be analyzed before the gas can enter the domestic transportation pipeline for compressed natural gas piping. In some embodiments, the surrogate approach is either the AIRGARD ™ system or the MultiGas 2030 (Mus Instruments, Inc.), both commercially available from MKS Instruments, Inc., Andover, Massachusetts. (Trademark) family.

FIG. 13 shows a simulation involving a 540 ppb mixture of L2-siloxane, L3-siloxane, L4-siloxane, D3-siloxane, D4-siloxane, D4-siloxane, and D5-siloxane other than methane that was also used for blending. Figure 3 shows the results for total siloxane concentration in ppm over time for a given landfill gas sample. The siloxane concentration is determined using the CLS analysis method in connection with the speciation method as shown in FIG. 9 or the surrogate method as shown in FIG. 12, which may or may not use correction factors. Also good. Measurements may be obtained using a MKS MG2030 ™ FTIR spectrometer commercially available from MKS Instruments, Inc. of Andover, Massachusetts. The spectrometer has a 5.11 m gas cell heated to 40 ° C. and 20 seconds of data is averaged up to 100 seconds. FIG. 13 shows a total siloxane graph 1310 determined using a speciation technique, where the second set of absorption spectra includes all known siloxane compounds and / or hydrocarbon compounds, and 2) the surrogate set of absorption spectra A total siloxane graph 1320 determined using a surrogate method that does not use a correction factor and includes only a subset of known siloxane compounds and / or hydrocarbon compounds, and 3) determined using the same surrogate method as graph 1320. And shows the total siloxane graph 1330 with the correction factor applied. For this synthesis biogas sample, the expected total siloxane peak amount should be 3.24 ppm-v (ie, 6 times 540 ppb for 6 siloxanes in the synthesis gas sample). Siloxane graph 1310 is generated by a speciation technique and has a maximum concentration of undiluted cylinder values at 3.11 ppm-v. In this case, the expected amount (3.24 ppm-v) deviates by about 4%, which is within an allowable error range. As shown, the total siloxane graph 1330 generated by the surrogate approach using the correction factor is as accurate as the siloxane graph 1310 obtained by speciation. In this particular example, the surrogate method may begin by dividing the total siloxane concentration around 50 ppb-v.

  FIG. 14 shows the results of total siloxane concentration over time for digestion gas samples diluted using 100% methane. The digestion gas sample is less than 200 ppb-v in approximately 60% methane and some ethane and propane and 40% carbon, and the approximate ratio of D4-siloxane to D5-siloxane is 75% to 25%. Contains siloxane compounds. Siloxane concentration may be determined using CLS analysis techniques. Measurements may be obtained using the MKS AIRGARD ™ system commercially available from MKS Instruments, Inc. of Andover, Massachusetts. The system has a 10.18 m gas cell heated to 40 ° C. and 20 seconds of data is averaged up to 100 seconds. FIG. 14 is a dilution factor based on plotting hypothesized changes in total siloxane values based on undiluted siloxane total in digested samples following CO2 concentration and its changes when adding methane to undiluted digestion gas. 2) the second set of absorption spectra is generated for all the main siloxane compounds (three cyclic siloxane components and three linear siloxane components) and the first of the hydrocarbon compounds. Total siloxane graph 1420 determined using a speciation technique, including subsets, 3) the second set of absorption spectra is all the main siloxane compounds (three cyclic siloxane components and three linear siloxane components) and Total siloxane graph 143 determined using speciation techniques, including a second subset of hydrocarbon compounds 4) a total siloxane graph 1440 determined using a surrogate approach, wherein the surrogate set of absorption spectra includes a first subset of siloxane compounds and a first subset of hydrocarbon compounds used in graph 1420; And 5) shows a total siloxane graph 1450 determined using the surrogate approach, where the surrogate set of the absorption spectrum includes the surrogate set of the siloxane compound and the second subset of the hydrocarbon compound used in graph 1430. As shown in FIG. 14, graphs 1440 and 1450 are each determined using a surrogate technique and track the change in siloxane concentration over time relative to the reference graph 1410. However, graphs 1420 and 1430 are each determined using a speciation technique and track the change in siloxane concentration over time in the digestion gas relative to baseline graph 1410. Graphs 1420 and 1430 are also more affected than either of the two surrogate methods represented by graphs 1440 and 1450 depending on which hydrocarbon is selected as the surrogate hydrocarbon. Accordingly, FIG. 14 shows that it is advantageous to determine the total siloxane concentration using a surrogate set (ie, a subset) of siloxane compounds, particularly when the siloxane concentration varies at low ppb-v.

  The systems and methods described above may be implemented in digital electronic circuitry, computer hardware, firmware, and / or software. Implementation may be performed as a computer program product (ie, the computer program is tangibly embodied in an information carrier). Implementation may be performed, for example, on a machine-readable storage device and / or a propagated signal that is executed by a data processing device or controls the operation of the data processing device. An implementation may be, for example, a programmable processor, a computer, and / or multiple computers.

  The computer program may be written in any form of programming language, including a compiler-type language and / or interpreted language, and may be a stand-alone program or subroutine, element, and / or other suitable for use in a computing environment It may be deployed in any form including introduction as a unit. The computer program may be deployed to be executed on one computer or multiple computers at one site.

  The method steps may be performed by one or more programmable processors executing a computer program to implement the functionality of the present invention by acting on input data and generating output. The method steps may be performed by dedicated logic circuitry and the device may be implemented as dedicated logic circuitry. The circuit may be, for example, an FPGA (Field Programmable Gate Array) and / or an ASIC (Application Specific Integrated Circuit). Modules, subroutines, and software agents may refer to portions of computer programs, processors, dedicated circuits, software, and / or hardware that implement functions.

  Processors suitable for executing computer programs include, by way of example, both general and special purpose microprocessors and one or more processors of any type of digital computer. Generally, a processor will receive instructions and data from a read-only memory or a random access memory or both. The basic elements of a computer are a processor for executing instructions and one or more memory devices for storing instructions and data. Generally, a computer may include one or more mass storage devices (eg, magnetic disks, magneto-optical disks, or optical disks) for storing data, and data from one or more mass storage devices. And / or operably coupled to transfer data to the mass storage device.

  The techniques described above may be implemented on a computer having a display device to allow user interaction. The display device may be, for example, a cathode ray tube (CRT) and / or a liquid crystal display (LCD) monitor. User interaction is, for example, with a keyboard and pointing device (eg, a mouse or trackball) that allows the user to display information and allow the user to make input to the computer (eg, interact with user interface elements). There may be. Other types of devices may be used to implement user interaction. The other device may be, for example, feedback provided to the user as any form of sensory feedback (eg, visual feedback, audio feedback, or haptic feedback). Input from the user may be received in any form including, for example, acoustic input, speech input, and / or haptic input.

The techniques described above may be implemented in a distributed computing system that includes a back-end component. The backend component may be, for example, a data server, a middleware component, and / or an application server. The techniques described above may be implemented in a distributed computing system that includes a front-end component. The front-end component may be, for example, a client computer having a graphical user interface, a web browser that allows the user to interact with the exemplary implementation, and / or other graphical user interfaces for the sending device. May be. The components of the system may be interconnected by any form or medium of digital data communication (eg, a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), the Internet, a wired network, and / or a wireless network.

  The system may include a client and a server. A client and server are generally remote from each other and typically interact through a communication network. The client-server relationship is caused by computer programs that execute on each computer and have a client-server relationship with each other.

  Although the invention has been particularly shown and described with reference to preferred embodiments thereof, those skilled in the art will recognize that various forms and details of the embodiments can be used without departing from the scope of the invention as encompassed by the appended claims. It will be understood that changes can be made.

Claims (10)

In a system for monitoring one or more silicon-containing compounds in a biogas,
A light source of a first radiation beam;
An interferometer that receives the first beam of radiation from the light source and forms a second beam of radiation comprising an interference signal;
A sample cell in optical communication with the interferometer;
A detector in optical communication with the sample cell,
A non-absorbing gas in the sample cell, the first interference signal propagating in the non-absorbing gas that does not substantially absorb infrared radiation in a specified wavelength range of interest; and in the sample cell A detector for receiving a second interference signal propagating in the sample gas comprising the biogas;
A processor in electrical connection with the detector and for calculating a total concentration of the one or more silicon-containing compounds in the biogas, the calculation comprising:
One or more selected from a first absorption spectrum based on a ratio of the first interference signal to the second interference signal and at least a larger set of known silicon-containing compounds having a known concentration A processor that calculates based on a set of surrogate absorption spectra based on individual absorption spectra for each silicon-containing compound in the subset of silicon-containing compounds.   The system of claim 1, wherein the one or more silicon-containing compounds in the biogas comprises at least one siloxane. The sample cell is
A concave reflective field mirror surface at a first end of the sample cell;
A substantially spherical concave reflective objective mirror surface facing the field mirror surface at a second end of the sample cell, the objective mirror surface comprising the field mirror surface and the objective mirror surface; A cylindrical component that increases the coincidence of each focal point in at least one plane so as to maximize the throughput of the second radiation beam propagating through the sample cell via a plurality of reflections at each of The system of claim 1.   The set of surrogate absorption spectra is selected from at least the individual absorption spectrum for each silicon-containing compound of the subset of one or more silicon-containing compounds and a larger set of known hydrocarbon compounds having a known concentration The system of claim 1, wherein the system is a model based on an individual absorption spectrum for each hydrocarbon compound in the subset of one or more of the hydrocarbon compounds being processed.   The total concentration is from one of a total concentration of siloxane compounds in the biogas, a total concentration of other silicon-containing compounds in the biogas, or a total concentration of all silicon-containing compounds in the biogas. The system of claim 1.   The system of claim 1, wherein the processor applies a correction factor to the total density that scales the total density by a factor.   The set of surrogate absorption spectra further includes an individual absorption spectrum for each hydrocarbon compound of a subset of one or more hydrocarbon compounds selected from a larger set of known hydrocarbon compounds having a known concentration. The system of claim 1, comprising: The set of surrogate absorption spectra is at least the individual absorption spectrum for each silicon-containing compound of the subset of one or more silicon-containing compounds and each hydrocarbon of the subset of one or more hydrocarbon compounds. The system of claim 1, which is a model based on the individual absorption spectra for a compound.   The system of claim 1, wherein the processor performs multiple regression analysis using the first absorption spectrum and the at least one surrogate absorption spectrum.   The processor uses a classical least squares (CLS) fitting routine based on at least one of the first absorption spectrum, the at least one surrogate absorption spectrum, and a baseline or offset value in the biogas. The system of claim 1, wherein the total concentration is calculated by representing spectral characteristics of the one or more silicon-containing compounds.