Field of the invention

The present invention relates to a chemical analysis method for the measurement of tetrafluoromethane, CF4, using optical means. More specifically it relates to measurement of CF4 with reduced influence of interfering gases that could be present in different applications.

Background of the invention

Tetrafluoromethane, CF4, can be emitted from different chemical processes. In particular, the production of aluminium can result in emissions of CF4. CF4 is a strong greenhouse gas said to be much stronger than C02. Few real time and in- situ instruments are capable of measuring CF4 and therefore measurement of CF4 is very often done using manual sampling consisting of sample bag or sorbent columns techniques.

The aluminum smelting process, also known as the Hall-Heroult process, is one of the largest anthropogenic contributors to atmospheric perfluorocarbons (PFCs). When the dissolved alumina content in the cryolitic bath falls below ~1 .5%, a rapid rise in the operating voltage occurs and the carbon anodes begin to react with the fluoride-based electrolyte (Na3AIF6) forming PFCs. These transient events are referred to as anode effects (AEs). Recent reports indicate that in addition to AEs, PFC generation may also occur during normal potroom activities that result in either (1 ) a localized increase in anode current density or (2) a localized deficiency in alumina concentration; PFCs emitted during these periods not associated with AEs have been referred to as low-voltage PFCs (LV-PFCs). Disclosure of the state of art

Fourier Transform Infrared (FTIR) spectrometers are commonly employed by the industry to measure PFCs in real time, providing an accurate snapshot of the emission profile during typical potroom operation. The technique relies on extractive sampling from the exhaust stream and subsequent conditioning of the sample gas prior to the measurement. The exhaust stream at aluminum smelters is rather hostile to FTIR spectrometers: the sample gases contain a high dust loading, chemically aggressive species (e.g. , hydrogen fluoride, HF), and spectrally interfering

compounds (e.g. , water vapor, H20). As such, deployment of an FTIR spectrometer for continuous monitoring of PFCs would require significant upkeep and daily maintenance. Instead, the industry relies on short-term in-plant campaigns to spectroscopically determine levels of PFCs emitted during a representative sampling of AEs spanning typical production cycles. Combining PFC emission values with production and AE metrics yields plant-specific slope terms used to estimate the long-term relationship between PFC emissions and smelter process parameters.

In-plant measurements are also performed in order to ensure regulatory compliance, at a frequency prescribed by the ruling government (e.g. , triennial for Quebecois smelters) or upon significant changes to the process control algorithm impacting AE metrics.

Tunable laser absorption spectroscopy (TLAS) is nowadays considered to be a mature technology for in-situ measurements allowing to perform year-round continuous emission measurements (CEMS) with a higher accuracy and faster response time than extractive methods while requiring significantly lower

maintenance. Frequently the terms Tuneable Diode Laser Absorption Spectroscopy (TDLAS), tuneable laser spectroscopy (TLS), and Tuneable Diode Laser

Spectroscopy (TDLS) are used instead of TLAS. Academic publication

"Gasmonitoring in the process industry using diode laser spectroscopy", Linnerud et.al. , Appl. Phys. B 67, 297-305 (1998) describes several aspects of gas monitoring based on second harmonic laser spectroscopy. United States patent application publication US2006/0044562 A1 describes several aspects of direct absorption spectroscopy as well as gas monitoring based on TLAS in general.

The aluminum industry is using TLAS sensors for many years in potrooms and exhaust stacks to perform in-situ measurements of HF. This success has led to the aluminum industry's interest in evaluating a similar technology for in-situ CF4 offering the capability to continuously monitor PFC emissions.

According to the HITRAN (high-resolution transmission molecular absorption database) and PNNL (vapor phase infrared spectral library) spectroscopic

databases, CF4 does not absorb in the near-infrared but displays a strong

absorption feature in the mid-infrared range located around 1283 cm-1 . A simulated transmission spectrum of 20 ppb CF4 in ambient air using an optical path length (OPL) of 1 m and ambient pressure is shown in Figure 1 . The simulated spectra were assembled using data available from the HITRAN and HITEMP (high- temperature molecular spectroscopic database) databases. A gas temperature of 125 °C was chosen since it is a typical value for emission monitoring applications. Under the described conditions, the absorption feature for CF4 is nearly 1 cm-1 broad. Interference from H20, CH4, and N20 is significant, whereas C02 does not absorb in this wavelength range.

First reports of mid-infrared TLAS sensors for in-situ measurements of PFCs employed lead-salt lasers. These sensors achieved detection limits in the low parts- per-billion (ppb) range with a few seconds acquisition time, the systems were employing multi-pass cells and extractive sampling. However, lead-salt lasers require cryogenic cooling and are in many cases unreliable; therefore they are not generally considered suitable light sources for sensors in permanent industrial installations. Recent developments in quantum cascade laser (QCL) technology have made compact and reliable mid-infrared laser sources available that are able to operate in continuous-wave (cw) mode at room temperature and even above. QCLs have proven to be feasible in several field applications; e.g., the measurement of nitric oxide (NO) in household waste incinerators or atmospheric sensing of nitrous oxide and carbon monoxide (CO).

Optical detection of the sensor can be based either on Direct Absorption

Spectroscopy (DAS) (figure 5) or on Wavelength Modulation Spectroscopy (WMS) (figure 6). The emission optical frequency of the laser source is in both cases typically adjusted to the center of the absorption line by setting the operation temperature. Applying a direct current and a current ramp is subsequently used to tune the optical frequency across the absorption line (DAS). The current ramp, and thus the emission wavelength of the laser, is optionally modulated to enable WMS- based detection.

The laser beam is passing through the gas sample and a detector is measuring the transmitted optical signal on the receiver side. In case of WMS, a hardware mixer extracts the second harmonic component within the preamplified detector signal. The DAS/WMS signal is sampled by an analogue-digital converter. Adjustable low-pass filters embedded in the signal-processing software are used to further improve the signal-to-noise ratio. As described in Geiser et al. [academic paper: Continuous Emission Monitoring of Tetrafluoromethane Using Quantum Cascade Lasers, Peter Geiser et.al., MDPI Photonics 2016, 3, 16 ] interference with other gases is occurring, especially

Methane (CH4), Water vapour (H2O) and Nitrous oxide (N2O), leading potentially to false CF4 concentration measurements; see table below. As can be seen, especially CH4 is causing a strong interference offset. For example, 1 ppm CH4 in the process gas mix lead to an offset of 10 ppb in the CF4 reading. CH4 is commonly present in the process at various concentration levels. The H2O content is also fluctuating due to different process and weather conditions. N2O are less of an issue since the typical concentration levels are very low.

Table 1:

The document "Protocol for Measurement of Tetrafluoromethane (CF4) and

Hexafluoroethane (C2F6) Emissions from Primary Aluminum Production" authored by the U.S. ENVIRONMENTAL PROTECTION AGENCY (US EPA), Washington, D.C. and INTERNATIONAL ALUMINIUM INSTITUTE, London, U.K. dated April 2008 gives in sections 6.2 and 6.3 an overview of different techniques as present in prior art.

For analysis in the lab gas chromatographs/mass spectrometers (GC/MS) can be used on gas sample bags or sorbent columns acquired in the field. FTIR

instruments can be used on a gas cell evacuated and filled with process gas.

For field measurement mass spectrometers, Tuneable Diode Laser Absorption Spectrometers, Photoacoustic Spectrometers and FTIR instrument are listed in section 6.2 of the document. In section 6.3 of the same document four technologies suitable for field

measurement are reviewed. For mass spectrometers, it is said that:

"The process mass spectrometer provides a near real time measurement of both CF4 and C2F6, as well as several other gas sample components if desired. " The further description follows:

" ... The instrument must be calibrated in place, prior to the start of PFC monitoring. The detection limit for PFC compounds is typically about 0. 1 ppm so there is not adequate sensitivity for direct measurement of fugitive PFC emissions and an alternative strategy must be used when measurements of fugitives are required. While the mass spectrometer has demonstrated good performance in the relatively strong magnetic fields common around smelter facilities, it is desirable to position the instrument in a location where magnetic fields are relatively invariant, since strong, fluctuating magnetic fields can affect the instrument's response. This equipment is relatively heavy and requires a mobile truck to transport to the measurement location in the facility."

A version of a Tuneable Diode Laser Absorption Spectrometer (TDLAS) has also been reviewed as shown in the following citation:

"TDLAS is an infrared absorption technique that uses a diode laser to achieve a very narrow emission source bandwidth. As a result, the specificity of the technique is very good. The sensitivity is also excellent and the instrument is capable of direct measurement of both PFC gas components in electrolytic cell exhaust ducts. It has not been applied to fugitive gas measurements. Consequently, if fugitives are to be measured, laboratory analysis of collected time average samples, or open path methodology using FTIR techniques would provide a viable strategy. The TDLAS unit as used for previous PFC measurements is relatively large. A mobile laboratory or trailer is needed to transport the instrument and ancillary sampling equipment from site to site. The equipment as used for past PFC measurements, is not broadly commercially available, is relatively expensive, is specialized and requires

experienced specialists to operate it. A newer, lower cost TDLAS instrument has been developed for the measurement of CF4, however, the instrument has not yet been field-tested. " The Photoacoustic Spectrometer (PAS) is described like this:

"The PAS is a filter type infrared spectrometer that uses a sensitive microphone as a detector to measure changes in absorption of infrared energy. The instrument is quite sensitive for the PFC gas components; however, interference can occur from any water, sulfur dioxide or methane accompanying the PFC gases. The sample must be conditioned to minimize or remove these potential interferences prior to measurement. Scrubbers containing sodium carbonate or Ascarite can remove hydrogen fluoride and sulfur dioxide. Water vapor is removed by passing the sample stream through a copper tube at dry ice temperature or by using a commercially available drying compound. The instrument calculation software also allows for compensation for limited quantities of interferences by measuring interfering compounds at alternate wavelengths. The PAS does not sample the gas stream continuously as do the other instrument methods described here. Instead the instrument operates in a sequential cyclic analysis mode with a new sample introduced to the analyzer detector on a frequency of once each three to five minutes. The sampling cycle is typically about 15 seconds of this total three to five minute cycle. Accommodation must be made for the substantial dead time of the basic instrument sampling system to prevent bias in results. Collection of time average samples in canisters or sample bags and subsequent analysis with the PAS is an effective approach to overcome the dead time limitation. The instrument is the most portable of the instruments described here weighing approximately 10 kg and is easily operated by staff with a basic knowledge of measurement science. "

Finally, part of the FTIR description is given below:

"... As with some of the other methods, potential problems from overlap of interfering spectral bands must be overcome through calibration procedures or spectral stripping algorithms. "

The document from the US EPA gives information on different technologies for the measurement of CF4 and it is clear that the lab measurements with manual sampling in the field and analysis in the lab is both time consuming and complex and not delivering results in real time as well as not being in-situ.

When it comes to equipment found to be suitable for field measurement the US EPA finds mass spectrometers and the TDLAS unit tested to be big and bulky and in need of a mobile truck or a mobile laboratory or trailer to transport instruments. The mass spectrometer has a limited detection limit and should not be installed in areas with fluctuating magnetic fields as this could influence measurement. The latter could be problematic in primary aluminium smelters where very high currents (150-450kA) through the reduction cells induce strong magnetic fields.

The PAS device tested has significant problems with interference and need sample preparation and removal of many substances. Additionally, the PAS device only samples (measures) 15 seconds of cycles of duration from 3 to 5 minutes which means that significant information of the gas concentration and process variation is lost due to incomplete coverage of time.

The US EPA further states that FTIR might have problems with interference that must be removed using calibration procedures or spectral stripping algorithms.

A product by the assignee of the current application is the LaserGas Q CF4 gas monitor by NEO Monitors AS of Norway. This gas monitor is designed for field use and is relatively light weight compared to most instrumentation mentioned in the US EPA document. It operates in-situ, is real time and on-line and has a short time response. However, the LaserGas Q CF4 operates in a region in the mid-IR on a CF4 line that is overlapping with lines from methane (CH4), water vapour (H20) and nitrous oxide (N20) and presence of these potential interfering gases lead to reduced accuracy of the CF4 reading.

Disclosure of the invention

Problems to be solved by the invention

The main objective with the current invention is to improve selectivity of a gas measurement method for the measurement of CF4 based on tuneable laser spectroscopy in presence of potential interfering gases like CH4, N20 and H20. It is an additional objective with the invention to be able to reduce the interference from these other gases also when the method is operated under normal ambient pressure as well as in-situ.

Means for solving the problems

The objectives are achieved according to the invention by a method for measuring CF4 gas concentration with reduced sensitivity to interference from other gases as defined in the preamble of claim 1 , having the features of the characterizing portion of claim 1 .

Summary of the invention

An aspect of the invention is a method for measuring CF4 gas providing a

measurement of CF4 gas concentration, comprising the steps of pointing a tuneable laser through a target gas, possibly comprising CF4 and other gases possibly comprising at least one gas causing interference for the measurement of CF4 gas concentration, onto a light sensitive detector, acquiring and digitising signals from the detector, wherein tuning the laser across a spectral absorption feature of CF4 (4100) around 1283 cm-1 . The method further comprises correcting the measurement of CF4 gas concentration based on data representing a concentration of the at least one gas causing interference for the measurement of CF4.

Optionally, the method can comprise measuring the data representing the

concentration of the at least one gas causing interference for the measurement of CF4 using the tuneable laser, and using the data for the correction of the

measurement of CF4 gas concentration.

Optionally, the method can comprise measuring the data representing the

concentration of the at least one gas causing interference for the measurement of CF4 with one or more additional lasers, and using the data for the correction of the measurement of CF4 gas concentration.

Optionally, the method can comprise measuring the data representing the

concentration of the at least one gas causing interference for the measurement of CF4 with at least one of the following techniques FTIR, NDIR, UV, capacitive sensors and electro-chemical sensors, and using the data for the correction of the measurement of CF4 gas concentration.

Optionally, the method can comprise measuring or estimating the data representing the concentration of the at least one gas causing interference for the measurement of CF4 in a preparatory step, storing the data in a table, and using the data for the correction of the measurement of CF4 gas concentration.

The at least one gas causing interference for the measurement of CF4 can be one of or any combination of methane (CH4), water vapour (H2O) and nitrous oxide (N2O).

Optionally, the method can comprise measuring the data representing the methane (CH4) (4200) concentration in the range from 1249.5 cm-1 to 1250.0 cm-1 .

Optionally, the method can comprise measuring the data representing the nitrous oxide (N2O) (4400) concentration in the range from 1274.2 to 1275.0 cm-1 . Optionally, the method can comprise measuring the data representing the water vapour (H2O) (4300) concentration in the range from 1265.7 to 1266.5 cm-1 .

At least one laser can be of a quantum cascade type or an interband cascade type.

Optionally, the method can comprise modulating at least one laser using direct absorption spectroscopy or wavelength modulation spectroscopy.

Optionally, the method can comprise basing the correction of the CF4 concentration on a look-up table technique.

The correction of the CF4 concentration can be based on subtraction of at least one absorption spectrum, the at least one absorption spectrum representing the spectrum of at least one gas causing interference for the measurement of CF4, subtracting this at least one other absorption spectrum after the detector signal has been acquired when the laser has been tuned across the spectral feature of CF4, the acquired detector signal representing an absorption spectrum of the target gas possibly comprising CF4 and other gases, subtracting the at least one absorption spectrum from the absorption spectrum of the target gas and finally determining the gas concentration of CF4 from the result of the subtraction of the spectra.

The correction of the CF4 concentration can be based on multi variate analysis, MVA.

Description of the figures

Embodiments of the present invention will now be described, by way of example only, with reference to the following diagrams wherein: Figure 1 shows the wavelength region around the CF4 absorption feature (4100) around 1283 cm-1 (7.794 urn) with curves for CF4 (4100), CH4 (4020), H2O (4030), N2O (4040) and CO2 (4050). The curves for each gas are plotted independently in the diagram. The CO2 absorption (4050) is so weak in this region that it overlaps with the upper y-axis. Wavenumber is shown on the X-axis and transmission is shown on the Y-axis.

Figure 1 as well as figures 2, 3 and 4 has the gas mixture which is typical for use in the aluminium smelter application. Figure 2 shows the region where an interference free methane (CH4) line (4200) will be selected according to a first preferred embodiment. The preferred absorption line for CH4 (4200) measurement is indicated with an arrow.

Figure 3 shows the region where an interference free water vapour (H20) line (4300) will be selected according to the first preferred embodiment. Please note that the water vapour absorption is significantly stronger than the absorption of the other gases and therefore the y-axis for water vapour (H20, right side) has a different range compared to the other (left) y-axis. The preferred absorption line for measurement of H20 (4300) is indicated with an arrow.

Figure 4 shows the region where an interference free nitrous oxide (N20) line (4400) will be selected according to the first preferred embodiment. The preferred absorption line for measurement of N20 (4400) is indicated with an arrow. Figure 5 shows several laser scan cycles for a gas analyser working with direct absorption technology. The laser current is shown. The figure shows a current ramp (1000), an optional dark reference time slot (1 100) and short time slot (1 150) where the laser current is on and where the laser current is constant to allow the laser to stabilise after the dark reference. Figure 5 is not to scale.

Figure 6 is similar to figure 5 but for wavelength modulation spectroscopy and second harmonic detection. A sine wave is added to the laser current whenever the laser is on. The laser current is shown. The figure shows a current ramp (1000), an optional dark reference time slot (1 100) and short time slot (1 150) where the laser current is on and where the laser current is constant to allow the laser to stabilise after the dark reference. Figure 6 is not to scale and is made to illustrate techniques. Description of reference signs

Description of preferred embodiments of the invention

Various aspects of the disclosure are described more fully hereinafter with reference to the accompanying drawings. This disclosure may, however, be embodied in many different forms and should not be construed as limited to any specific structure or function presented throughout this disclosure. Rather, these aspects are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art. Based on the teachings herein one skilled in the art should appreciate that the scope of the disclosure is intended to cover any aspect of the disclosure disclosed herein, whether implemented independently of or combined with any other aspect of the disclosure. For example, an apparatus may be implemented or a method may be practiced using any number of the aspects set forth herein. In addition, the scope of the disclosure is intended to cover such an apparatus or method which is practiced using other structure, functionality, or structure and functionality in addition to or other than the various aspects of the disclosure set forth herein. It should be understood that any aspect of the disclosure disclosed herein may be embodied by one or more elements of a claim. The current invention is about a method for correcting measurement of tetrafluoromethane (CF4) concentration in the presence of gases that leads to interference for the CF4 concentration measurements. Three of the gases that can lead to interference are methane (CH4), water vapour (H20) and nitrous oxide (N20). The presence of these interfering gases leads to reduced accuracy of the CF4 concentration measurement.

The method comprises the steps of pointing a tuneable laser through a target gas onto a light sensitive detector and acquiring and digitising signals from the light sensitive detector while tuning the laser across a spectral absorption feature of CF4 (4100). The concentration of CF4 will be calculated from the acquired and digitised detector signal.

The tuning of the laser and the concentration measurement can be done using direct absorption spectroscopy (DAS) (figure 5) or wavelength modulation spectroscopy (WMS) (figure 6).

The target gas can possibly contain CF4 depending on process and process conditions. The target gas can possibly also contain at least one gas causing interference for the CF4 concentration measurement, the presence and

concentration of this at least one gas causing interference for the CF4 concentration measurement will also depend on process and process conditions. Other gases not causing interference for the CF4 concentration measurement will typically also be present in the target gas.

The spectral feature (4100) for CF4 that has been selected for the method according to the current invention is located around 1283 cm-1 .This is the strongest CF4 feature and the only one that gives a sufficiently good (low) detection limit. However, this region also contains spectral features for other gases normally present in industrial processes including processes in the aluminium industry. The presence of these other gases with spectral features in the same spectral region will lead to interference which will reduce the accuracy of the measurement of the CF4 concentration. Therefore, in more detail, this invention is about how to improve the selectivity of the CF4 concentration measurement in the presence of interfering gases, thereby increasing the accuracy of the CF4 concentration measurement and at the same time maintaining or improving the sensitivity and the good (low) detection limits for the CF4 concentration measurements.

The data acquired and digitised from the detector signal when the laser wavelength is scanned or tuned across a wavelength region comprising a feature (4100) of CF4 represents the absorption spectrum of the target gas. The actual absorption will depend on the optical path length and the concentration of the gases present. The target gas can possibly contain CF4 and one or more interfering gases as well as other gases. The acquired and digitised data from the detector signal will be used to calculate the concentration of CF4 gas. The interfering gases will have spectral features (4020, 4030, 4040) in the same wavelength region as CF4, more or less on top of the CF4 spectral features (4100). Depending on the concentrations of the interfering gases, the spectral contribution of the interfering gases will lead to changes in the absorption spectrum of the target gas and as a result lead to changes or errors in the measured CF4 concentration calculated from the acquired data in the absorption spectrum of the target gas.

A main objective with the method according to the current invention is to correct for the presence of interfering gases and in particular correct for the presence of CH4, H20 and N20.

If the concentrations of the interfering gases are known, it is possible to correct for their presence and their contribution to the error in the measured concentration of CF4. It is part of the method according to the current invention to correct for the presence of interfering gases based on the knowledge of the concentration of the interfering gases.

Correction of the measured CF4 concentration

We will first assume that we have the concentrations of the interfering gases or other data representing the presence and concentrations of the interfering gases available from a preparatory, previous or in general another step of the method according to the current invention.

One possible solution is to prepare a look-up-table in a preparatory step to the method. This look-up-table will have inputs or indexes for each interfering gas we want to correct for, at least one interfering gas. If we want to correct for 3 interfering gases it will be a 3-dimensional look-up-table with an x, y and z input index, x, y and z representing a concentration level for one interfering gas each. Index values will be in steps, integer steps, representing a concentration range of each interfering gas. One possible implementation is to use the closest index range to the actual concentration or one can interpolate between output values from the look-up-table for the index above and below the actual concentration. The content of the look-up- table will be compiled in a preparatory step of the method based on actual measurements in the lab or in the field where the concentration of the different gases could be varied. Alternatively the compilation of the look-up-table contents could be generated based on simulation of spectra based on data from databases like

HITRAN. It might be possible also to include look-up-table input indexes also for target gas temperature and pressure.

The content of the look-up-table (its output) will be a result of the input (indexes) which is the concentration of the interfering gases. The output from the look-up-table will be a correction value that will be added to or subtracted from the measured CF4 concentration.

Another approach for correction of the influence from the interfering gases to the measured CF4 concentration is to generate spectra that can be subtracted from the absorption spectra acquired and digitised from the detector signal when the laser has been scanned across the wavelength region of the selected CF4 spectral feature around 1283 cm-1 .

Based on the actual concentrations used for the correction of each interfering gas it is possible to generate a simulated spectrum representing each interfering gas and then subtract them one by one from the spectrum of the target gas. Alternatively it is possible to make a composite spectrum with contribution from all selected interfering gases and their selected concentration values. Then this composite spectrum will be subtracted from the spectrum of the target gas. The resulting spectrum will then be used for the measurement or calculation of the corrected CF4 concentration.

Alternatively the spectrum or spectra generated based on the concentrations of the interfering gases could be adapted to be used either to multiply or divide the acquired and digitised spectrum from the target gas sample point by sample point. A third approach for correction of the influence from the interfering gases to the measured CF4 concentration is to use techniques based on multi variate analysis, MVA. The results from the MVA calculations will be used to determine the corrected concentration of CF4 and optionally the concentration of the interfering gases if the concentration of the interfering gases was not used as an input.

The inputs to the MVA analysis will be the spectrum of the target gas, that is, the acquired and digitised signal from the light sensitive detector either only covering a scan across the CF4 feature or covering a scan made continuously or semi- continuously using the tuneable laser covering the region of the CF4 feature (4100) and additionally covering a region comprising a feature (4200, 4300, 4400) of one or more interfering gases, the region with the feature of the interfering gas free of interference from then other gases.

In case one or more interfering gases are measured using another tuneable laser, similarly the spectrum, that is, the data acquired and digitised from the

corresponding other light sensitive detector is used as input to the MVA analysis in addition to the data from the region comprising the CF4 feature (4100).

In case the concentration for at least one interfering gas is constant, is measured using another technology than tuneable laser spectroscopy or is measured using another tuneable laser, the determined concentration value for this interfering gas can be fed as input into the MVA analysis in addition to the acquired and digitised data from the region comprising the CF4 feature.

Determining the concentration of interfering pases

For the different approaches presented above, it has been assumed that the concentrations of the interfering gases have been known from a previous or a preparatory step of the method according to the current invention. There are several options for how to determine the concentration level of the interfering gases.

The simplest way to handle the concentrations of the interfering gases is to determine their levels in a preparatory step of the method according to the current invention, store them as constants and then use them as constants during other phases of the method. In cases where the process conditions do not change or where there are only minor variations for the concentration of one or more of the interfering gases, the deviations caused by assuming the concentration level for one or more of these interfering gases are constant will be low. Depending on the process variations, temperature, pressure and variations in concentration of interfering gases it might be possible to use constant values for the concentration of one or more interfering gases and still achieve a good effect from the correction. A preparatory step for determining the concentration levels of the interfering gases could be based on measurement of the levels of the concentration of the interfering gases as well as on their variations. The conclusion from such measurements could be that one or more interfering gases vary very little in concentration level and could be treated as constant possibly the year round or could be treated as constants within one season of the year where the ambient temperature and/or humidity is stable. Another conclusion could be that the concentration varies so much that the concentration of the interfering gases must be measured continuously to achieve sufficient correction of the measured CF4 concentration. The concentration of one or more of the interfering gases can be measured by the tuneable laser, the tuneable laser that scans across the CF4 spectroscopic feature (4100). This can be done extending the scan of the tuneable laser so that spectroscopic features (4200, 4300, 4400) of one or more of the interfering gases will be covered by a continuous scan. Alternatively it could be a semi-continuous scan where the tuneable laser first is scanned across the CF4 feature (4100) and then after some laser settings are changed the feature(s) (4200, 4300, 4400) of at least one interfering gas is scanned. If there are more than one feature of one or more interfering gases to be scanned it might be required to change laser settings between the scan of each feature (4200, 4300, 4400) of the one or more interfering gases. The question if it is possible to scan also one or more interfering gases with the tuneable laser (the laser scanning across the CF4 feature (4100)) will be determined by the tuning properties of the actual tuneable laser being used, the tuning range must be sufficiently large. In cases where it is not possible to use the tuneable laser to scan across all features (4200, 4300, 4400) required to measure all the needed interfering gases, it will be possible to introduce one or more additional tuneable lasers for measurement of one or more of the interfering gases. This could be normal tuneable laser spectroscopy based analysers either using direct absorption spectroscopy (DAS) (figure 5) or wavelength modulation spectroscopy (WMS) (figure 6) possibly integrated with the tuneable laser and light sensitive detector used in the method according to the current invention in a similar fashion as described in patent publication WO

2016/200274 A1 (PCT/NO2016/050121 ). This patent publication describes an optical system suitable for integration of several tuneable lasers into one single instrument measuring multiple gases using multiple lasers. The use of the one or more tuneable lasers for measurement of the interfering gases will result in a concentration for the one or more interfering gases. Alternatively it will result in acquired and digitised data that can be used by MVA analysis.

It is also possible to measure the concentration of one or more interfering gases using other techniques than tuneable laser spectroscopy. Possible techniques for measurement of the concentration of one or more interfering gases could be, but are not limited to, FTIR, NDIR, UV, capacitive sensors or electro-chemical sensors.

The CF4 concentration will be measured using the tuneable laser. The data used for the correction of the measured CF4 concentration to correct for the interference from one or more interfering gases could be determined from one of or a combination of the following sources:

A constant concentration value determined in a preparatory step of the method

A concentration value as a result of calculation on data acquired when the tuneable laser was scanned across spectral features (4200, 4300, 4400) of an interfering gas in an extended scan of the laser

A concentration value as a result of using another tuneable laser and scanning this across features (4200, 4300, 4400) of one or more interfering gases A concentration value as a result of using a gas analyser or a gas sensor utilising another technique than tuneable laser spectroscopy Acquired data from an extended scan of a feature (4200, 4300, 4400) of one or more interfering gases using the tuneable laser

Acquired data from an additional tuneable laser that has scanned a feature (4200, 4300, 4400) of one or more interfering gases

There are no sufficiently strong absorption lines or bands of CF4 other than the range around 1283 cm-1 (7.79 urn) that are available and that allow measurements with the required sensitivity. The other absorption lines or bands are much weaker. Thus, the above wavenumber (wavelength) region for CF4 has to be used. To solve the interference issue, a new method to remove the influence of other gases on the measured concentration of CF4 can be applied. The basic spectroscopic

measurement principle is identical to before using tuneable laser spectroscopy (TLS) utilising either direct absorption spectroscopy (DAS) (figure 5) or wavelength modulation spectroscopy (WMS) (figure 6).

The basic idea is to measure in a first step the concentration values of CH4, and optionally H20 and N20 by means of TLS or another analytical method. In case TLS is used, the concentration measurements can be performed either by a single widely tuneable laser source (not necessarily continuous tuning) or by using two or more individual lasers. The preferred range for CH4 detection is between 1249.5 cm-1 and 1250.0 cm-1 , but also other wavenumber ranges are possible. For H20, the preferred range is between 1265.7 and 1266.5 cm-1 ; for N20 it is between 1274.2 and 1275.0 cm-1 . A CF4 spectrum is recorded as well. The order of acquisition is not important.

The concentrations of CH4, H20 and N20 can be acquired also by other means (FTIR, NDIR, UV etc.) and fed into the CF4 analyser. The concentration information obtained (either only CH4, or additionally H20 and/or N20) are used as input to the signal processing of the CF4 spectrum. The

concentration information can be used either (1 ) to subtract predefined values proportional to the concentrations of the interfering gases (using a table as listed above), or (2) to prepare a background spectrum that is subtracted from the CF4 spectrum, or (3) is used as boundary conditions for advanced signal processing algorithms like Multivariate Analysis or similar methods. The above procedure is applicable to DAS and WMS. In case (1 ), a look-up table must be generated and implemented into the signal- processing implementation prior to the actual CF4 measurements. This table includes information on the influence of the respective interference gas(es) on the CF4 concentration. These values are typically normalized to interference per 1 % or 1 ppm of the gas. Tables for different temperatures and pressures must be stored.

See table 1 above. For example: In case 3 ppm CH4 are in the gas mix, the interference will be:

10 ppb CF4 / 1 ppm CH4 * 3 ppm CH4 = 30 ppb CF4.

If more than one interference gas is present, all values have to be subtracted from or added to the CF4 concentration value obtained in the analysis of the absorption spectrum.

For example: The analysis of the spectrum leads to 100 ppb CF4 and 3 ppm CH4 are present, the real CF4 value is 70 ppb. In case (2), spectra of the interference gas(es) must be recorded and stored in the signal-processing implementation. The gases can be either pure, i.e. the gas mix contains only the interference gas and nitrogen (e.g. CH4 in N2) or a mixture of different gases (e.g. H20, C02, CH4). Spectra for different concentrations

(combinations), temperatures and pressures must be stored.

The recorded spectra during the application will be corrected for the interference by performing a point-to-point subtraction with a scaled interference spectrum.

For example: An interference spectrum of CH4 was recorded at a given temperature and pressure for a concentration of 1 ppm; stored in the instrument. In case of 3 ppm CH4 in the process, this spectrum will be multiplied (scaled) by 3 and pointwise subtracted from the spectrum recorded in the field measurement.

In case (3), the concentration values of the respective interference gases will be used as start values in advanced signal-processing algorithms (like MVA). The concentration values will be boundary conditions set to a constant value. Usually, the algorithm is trying to perform an optimal fit of the recorded process spectrum with all degrees of freedom. The algorithm is in general able to create concentration values for all gases present in the spectrum. The knowledge of one or more gases prior to the fitting reduces thus the degree of freedom and will optimize the fitting of the CF4 feature (4100).

The uniqueness of this method lies in the fact that it can be applied to in-situ measurements (directly in the process), without prior sample extraction. Scrubbing or other processes to clean an extracted sample is not necessary. It also avoids other used methods to remove or limit interference by reducing the pressure in a sample cell and thus reducing the overlap of spectral absorption lines. However, while the preferred setup is for in-situ measurements, the same method can be applied for extractive systems as well. In this case, a gas sample is extracted to a single- or multi-pass cell which comprises then the gas sample. The consecutive steps are identical to the ones in in-situ measurements.

Figure 1 shows the wavelength region around the CF4 absorption feature (4100) around 1283 cm-1 (7.794 urn) with curves for CF4 (4100), CH4 (4020), H2O (4030), N2O (4040) and CO2 (4050). The curves for each gas are plotted independently in the diagram. The CO2 absorption (4050) is so weak in this region that it overlaps with the upper y-axis. Wavenumber is shown on the X-axis and transmission is shown on the Y-axis. Figure 1 as well as figures 2, 3 and 4 has the gas mixture or "gas matrix" which is typical for use in the aluminium smelter application. The temperature is 125°C, the pressure is atmospheric, the optical path length, OPL, of the example is 1 m and the concentration of the gases is:

The wavenumber range plotted in figure 1 is 1282-1284 cm-1 . The CO2 absorption is weak in this region and the curve for the CO2 absorption (4050) overlaps the upper x-axis. This applies for figures 1 , 2, 3 and 4. The figures 1 , 2, 3 and 4 covers part of the wavenumber region 1240 to1284 cm-1 . The region is free of C02 interference.

Figure 2 shows the region where an interference free methane (CH4) line (4200) will be selected according to a first preferred embodiment. Wavenumber is shown on the X-axis and transmission is shown on the Y-axis. The wavenumber range in figure 2 is 1240 to 1260 cm-1 . The preferred absorption line for CH4 (4200) measurement is indicated with an arrow. Figure 3 shows the region where an interference free water vapour (H20) line (4300) will be selected according to the first preferred embodiment. Wavenumber is shown on the X-axis and transmission is shown on the Y-axis. The wavenumber range for figure 3 is from 1264 to 1268 cm-1 . Please note that the water vapour absorption is significantly stronger than the absorption of the other gases and therefore the y-axis for water vapour (H20, right side) has a different range compared to the other (left) y-axis. The preferred absorption line for measurement of H20 (4300) is indicated with an arrow.

Figure 4 shows the region where an interference free nitrous oxide (N20) line (4400) will be selected according to the first preferred embodiment. Wavenumber is shown on the X-axis and transmission is shown on the Y-axis. The wavenumber range for figure 4 is from 1273 to 1277 cm-1 . The preferred absorption line for measurement of N20 (4400) is indicated with an arrow. Figure 5 shows several laser scan cycles for a gas analyser working with direct absorption technology. The laser current is shown. A current ramp (1000) scans the wavelength of the laser across at least one spectral absorption feature for a target gas to be measured. An optional dark reference (1 100) time slot follows where the laser current is off. A short time slot (1 150) where the laser current is on and where the laser current is constant to allow the laser to stabilise after the dark reference follows. Then a new laser scan ramp is performed for the next cycle. Figure 5 is not to scale.

Figure 6 is similar to figure 5 but for wavelength modulation spectroscopy and second harmonic detection. A sine wave is added to the laser current whenever the laser is on. The laser current is shown. A current ramp (1000) scans the

wavelength of the laser across at least one spectral absorption feature for a target gas to be measured. An optional dark reference (1 100) time slot follows where the laser current is off. A short time slot (1 150) where the laser current is on and where the laser current is constant to allow the laser to stabilise after the dark reference follows. Then a new laser scan ramp is performed for the next cycle. Figure 6 is not to scale and is made to illustrate techniques.

There are different possible implementations of the method according to the present invention. They are listed below starting with the most preferable one. The method can be applied to in-situ measurements as well as extractive measurements. The embodiments can be performed at different process temperatures and pressures.

The setup is based on standard TLS: The emitted light of a laser source is guided using a proper optical arrangement through a gas sample, "the target gas", (in-situ or sample cell) and focused on a detector.

A widely tuneable mid-infrared laser is used, either with continuous tuning over the entire required spectral range or discrete tuning with sections where continuous tuning is possible to measure the required absorption lines. The temperature of the laser is stabilized to the desired operational point. Spectra of CF4 and interferents (CH4, H20, and N20, but also other gases are possible) will be recorded. Not all interferents must be measured, also a subset is possible. The order of importance is CH4, H20, N20.

The order of measurements is not of importance as long as the signal processing is done after acquiring all necessary spectra. Otherwise, the interferents have to be measured and processed before recording and analysing the CF4 spectrum.

The preferred range for CF4 detection is between 1283.0 cm-1 and 1284.0 cm-1 . For CH4 it is between 1249.5 cm-1 and 1250.0 cm-1 . For H2O, the preferred range is between 1265.7 and 1266.5 cm-1 ; for N2O it is between 1274.2 and 1275.0 cm-1 . These ranges are not strict and can vary (more or less tuning). Also other lines of CH4, H2O and N2O can be used.

Recording of spectra is identical for all gases and will only be described once. Proper currents are applied to operate the laser above lasing threshold and around the centre wavelength of the absorption line that should be measured. A current ramp (1000) is used to tune the wavelength over a defined range to scan the absorption line (Direct Absorption Spectroscopy, DAS) (figure 5). Optionally, a sinusoidal waveform can be added to the ramp current to perform Wavelength Modulation Spectroscopy (WMS) (figure 6). The detector signal is sampled. In case of WMS, a mixer is extracting the second harmonic component.

In a first step, the concentrations of interferents are determined applying typical DAS or WMS algorithms (including Multivariate Analysis).

The concentration information obtained (either only CH4, or additionally H2O and/or N2O) are used as input to the signal processing of the CF4 spectrum. The concentration information can be used either (1 ) to subtract predefined values proportional to the concentrations of the interferents (using an internal reference table in the analyser), or (2) to prepare a background spectrum that is subtracted from the CF4 spectrum, or (3) is used as boundary conditions for advanced signal processing algorithms like Multivariate Analysis or similar methods. The latter is the preferred method.

The above described method can also be applied with several laser sources in the near- and mid-infrared spectral region. In this case, the concentration information of the interferents must be fed into the CF4 analyser using proper communication interfaces. The determination of concentrations (operation of each laser) is identical to above described procedure.

Finally, above described method can also be applied using other than TLS techniques to determine the concentrations of interferents. In any case, the interferents concentration information must be fed into the CF4 analyser as described above.

Overview and summary of embodiments

A first preferred embodiment scans across the CF4 feature (4100) and measures the CF4 concentration, scans also across absorption features (4200, 4300, 4400) for 3 interfering gases using the tuneable laser to correct for the interference these 3 gases, CH4, H2O and N2O, impose on the measured CF4 concentration.

Variations of this first preferred embodiment will include only to measure one interfering gas or two interfering gases and to compensate for interference from this interfering gas or these interfering gases. Another set of variations of this first preferred embodiment is to measure also other potential interfering gases than CH4, H20 and N20 in other gas measurement applications than the one mainly targeted by this invention. In principle this will include to compensate for any number of potentially interfering gases that practically could be handled.

A second preferred embodiment is to measure the CF4 concentration and either no or at least one interfering gas with the tuneable laser and to measure at least one other interfering gas with a separate laser. The merging of the two laser beams could be performed as described in international patent application number

PCT/NO2016/050121 with international publication number WO 2016/200274 A1 . This patent publication describes an optical system suitable for integration of several tuneable lasers into one single instrument measuring multiple gases using multiple lasers.

A third preferred embodiment is to measure the CF4 concentration with the tuneable laser and to measure at least one interfering gas using another technique than tuneable laser spectroscopy. The tuneable laser could be used to measure CF4 and either no or a number of interfering gases while at least one other interfering gas could be measured using another technique different from laser spectroscopy.

A fourth preferred embodiment is to combine the second and the third preferred embodiments and to measure CF4 and either no or at least one first interfering gas with the tuneable laser and to measure at least one second interfering gas with a separate laser different from the tuneable laser and to measure at least one third interfering gas with a technique different from tuneable laser spectroscopy.