METHOD FOR EXAMINING PROCESS FLOWS

Description

[0001] The present invention relates to a method for analyzing process streams whose composition changes over short time intervals. The method according to the invention is flexible and versatile. A preferred application of the method concerns catalytic processes with product streams containing a large number of components, in particular with five or more different hydrocarbon-containing components.

[0002] The catalytic processes of interest here are characterized by the fact that they are carried out in industrial production facilities and in flow reactors. In numerous processes, complex reaction networks form in the catalyst beds of the reactors. The term "complex reaction networks" in the context of the present invention means that the reactions taking place in the reactors can no longer be described by simple chemical reaction mechanisms, for example, first-order kinetics, and that a large number of reactions take place simultaneously in the catalyst bed. As a result, process streams are obtained that have a composition with a large number of different components.

[0003] The complexity associated with characterizing these process streams also results from the fact that the compositions of the process streams can change within comparatively short periods of hours, minutes, and seconds. Causes of these changes include, for example, deactivation or run-in behavior of the catalysts, changes in the composition of reactants, or changes in process conditions.

[0004] Online spectroscopic methods for characterizing complex sample mixtures using chemometric techniques in industrial production processes are known in principle. A brief overview of the state of the art is provided below.

[0005] In the petrochemical industry and in fuel blending, spectroscopic analysis methods have been used for several decades to determine the composition of process streams. The fuels can contain several hundred different compounds, and the target composition of the fuels must be within a very narrow range to deliver the desired anti-knock properties. Since refineries produce large quantities of fuels, spectroscopic methods are particularly suitable for determining the chemical composition of the streams at short intervals. An example for the blending of oxygenate-containing hydrocarbons is US Pat. No. 5,596,196, which discloses that oxygenate-containing hydrocarbons can be determined with high accuracy using Raman spectroscopy and multivariate analysis.The resulting measurement signals can be used to monitor the concentration of components in products. This method is carried out directly in the liquid phase of the process stream using fiber optics and can deliver measurement results in less than one minute.

[0006] PCT application WO 01 / 02088 A1 describes a method for controlling the production process of polyhydroxy alcohols. In this method, the composition of the liquid process stream is continuously monitored using IR spectroscopy. The difference between the measured and desired composition is then determined to control the production process parameters. The process is a multi-step procedure. Measurements can be performed at different points in the process stream. The absorbance of the process stream is measured using an IR spectrometer, and then a computer determines the composition of the process stream in accordance with a preprogrammed calibration model.

[0007] In the context of blending liquid process streams, U.S. Patent No. 5,349,188 from Ashland Oil Inc. is also worth mentioning. It deals with the NIR analysis of PIANO constituents and the determination of the octane number of hydrocarbons. The PIANO constituents include the groups paraffins, isoparaffins, aromatics, naphthenes, and olefins. The method measures the absorption bands characteristic of the individual PIANO constituent groups.

[0008] Furthermore, Ashland Oil Inc. disclosed in WO 99 / 02973 a process / apparatus for the analysis of hydrocarbons using NIR, whereby concentrations of individual components in the aromatic range (BTX) are correlated with specific absorption bands in the NIR spectrum using statistical methods, thus enabling measurement and process control in the liquid phase in refinery processes.

[0009] US Pat. No. 5,360,972 from Western Atlas International Inc. discloses the improvement of IR spectroscopic measurements on materials by analyzing the spectroscopic data with calibration models. Various mathematical calibration models are described. In connection with the conduct of the tests, it is described that measurements are taken on standard samples using a NEG calibration set. The standard samples are liquid samples that were measured with a GC instrument that operates independently of a process stream.

[0010] In summary, the current state of the art prefers working in the liquid phase and in large-scale plants. For the calibration of the spectroscopic method (MIR, NIR, Raman), samples are analyzed offline, i.e., outside the actual reaction process and test conditions, using gas or liquid chromatography and then also measured spectroscopically. Based on this offline calibration, a model is developed using statistical methods, and the spectroscopic analysis is subsequently carried out online, i.e., under the actual reaction conditions, in the process stream.

[0011] US 2007 / 0082407 A1 by Joseph P. Little describes and claims the performance of an online analysis of a process stream based on the use of an optical analyzer. The optical analyzer comprises a number of optical sensors that can be used to perform spectroscopic analyses. The spectroscopic analyses are used to determine the chemical composition of natural gas. The sensors are integrated into a data collection system to determine the composition of natural gas at different locations within a storage and transmission infrastructure in a nearly continuous manner.

[0012] JPH0232253 by Shimadzu Corp. describes a device that enables data processing, whereby the detector signals from the chromatograph and spectrometer are subject to real-time data processing.

[0013] US 2015 / 0260695 A1 by Prism Analytical Tech. Inc. describes and claims a measuring device consisting of a combination of a gas chromatograph and an FTIR spectrometer. The FTIR spectrometer is connected downstream of the gas chromatograph's separation column. When using the device, the sample is first separated into its individual components, and the individual components are then fed into an IR measuring cell.

[0014] WO 2017 / 075140 A1 also originates from the same applicant as US 2015 / 0260695 A1, namely Prism Analytical Tech. Inc., and describes and claims a measuring device that combines gas chromatography with FTIR spectroscopic analysis.

[0015] US 2002 / 0031737 A1 from American Air Liquid Inc. describes a method and apparatus for using a tunable laser diode to examine specific components in a gas flow from a process line, with the investigations being carried out in real time on combustion processes. The investigations relate to the species O 2 , CO, and H 2 O. The range of application is thus limited to a few components.

[0016] In a preferred embodiment, the method according to the invention is used in the field of high-throughput research. High-throughput research concerns the acceleration and parallelization of research experiments. The improvement of catalytic processes and catalysts through high-throughput research is of particular importance, as such improvements contribute to saving energy and / or conserving resources.

[0017] Examples of state-of-the-art technology in the field of high-throughput research include US application US 2002 / 0182735 A1 by Kibby et al. and PCT application WO 2016 / 110408 A1 by Kirchmann et al. US 2002 / 0182735 A1 describes the use of microreactors in combinatorial chemistry. The products resulting from the reactions can be analyzed and tested for various properties, including octane and cetane numbers. After the reaction has taken place, the reaction products can be individually transferred from a channel to an analytical instrument, particularly a chromatograph.

[0018] PCT application WO 2016 / 110408 A1 by Kirchmann et al., which is from the same applicant as the present application, relates to an apparatus and a method for investigating naphtha reforming processes. The apparatus comprises a plurality of reactors arranged in parallel. The reactor outlet lines are connected to an online gas chromatograph. Based on the gas chromatographic analysis of the product fluid streams, the reaction processes taking place in the individual parallel reactors are optimized.

[0019] An object of the invention is to provide a method for investigating process streams, in particular process streams whose composition can change over time, especially even over short periods of time. By means of the method according to the invention, it should also be possible to monitor changes in the composition of the process streams that occur at the short time intervals mentioned above. A further object is to provide a method that can preferably be used in high-throughput research, wherein a plurality of processes carried out in a parallel arrangement can be investigated by means of the method. Furthermore, a method for investigating catalysts, in particular for parallel investigations of catalysts under reaction conditions that are as close to practical ones as possible, is to be provided.

[0020] These and other objects are achieved by a method for investigating at least one process stream comprising at least five different hydrocarbon-containing components, the method comprising at least the following steps: a) providing at least one process stream line (35) which is operatively connected to at least one online IR spectrometer (2) and also operatively connected to at least one online gas chromatograph (1), b) passing at least one process stream through the at least one process stream line (35), wherein during this passing of the process stream through the process stream line (35) an analytical characterization of the process stream is carried out by means of the online IR spectrometer (2) and the online gas chromatograph (1).

[0021] Preferably, step b) takes place within a time window that is less than 1 hour, preferably less than 40 minutes.

[0022] The process according to the invention combines the advantages of chromatography (high level of detail) and spectroscopy (speed). Both are used online, i.e., during the catalytic conversion, either in parallel or sequentially with a known time offset, directly in the product stream. An advantageous aspect of the process according to the invention is the use of an online gas chromatographic method, since online gas chromatography enables the separation of the reaction products into their individual chemical compounds and provides a detailed qualitative and quantitative analysis. The gas chromatographic analysis provides the relevant information on the chemical composition, based on individual components and the amount of the individual components. Based on this information, the target parameters can be derived.

[0023] These target parameters are preferably selected from the group of conversion, octane number, selectivity.

[0024] However, chromatography also has the property that it delivers these results with a time delay, which can be in the range of 20 to 180 minutes, especially for complex process streams, and may be too slow to detect rapid changes in the product streams.

[0025] In contrast, spectroscopic methods provide information in the form of absorption as a function of wavenumber, with each spectrum yielding characteristic bands associated with a specific substance concentration and time. The advantage of spectroscopic methods is that the results are available virtually simultaneously, thus enabling almost continuous analysis. The measurement time of online spectroscopic analysis can be in the range of seconds. Thus, the method according to the invention enables a temporal resolution in the range of seconds.

[0026] A spectroscopic method can thus be used for rapid control, whereby when a target parameter is reached and measurement data with high information density are recorded, chromatographic methods are used again to obtain more details.

[0027] Typically, the use of a spectroscopic analytical method involves significant calibration effort. In particular, the investigation of complex process streams requires calibration using a chemometric model.

[0028] An advantage of the process according to the invention is that the process can be carried out under process conditions that correspond to the process conditions of real process streams (i.e. reaction conditions that actually occur when carrying out the reaction on an industrial scale).

[0029] Preferably, the method according to the invention is used to investigate processes which occur in a time window of 5 seconds - 30 days, more preferably in the time window of 30 seconds - 14

[0030] Days, furthermore preferably 60 sec - 7 days, and which can no longer be monitored with chromatographic methods otherwise commonly used on a pilot plant scale, or where the use of control loops for readjustment can no longer be carried out quickly enough to monitor changes in the product spectrum.

[0031] However, it is also possible to use the method according to the invention in connection with the investigation of process streams that exist over longer periods than 30 days, for example, the method can also be carried out in connection with process streams where the time window is in the range of 2 - 24 months.

[0032] The method according to the invention offers the advantage that the GC measurement method and the IR spectroscopic measurement method are carried out in situ, i.e., directly in the process stream. Preferably, the calibration of the GC measurement method and the calibration of the IR spectroscopic measurement method, as well as the correlation of both measurement methods, are carried out on the basis of the in situ measurements. Furthermore, the method is preferably operated as a self-learning system. In this preferred embodiment of the self-learning system, the method is characterized in that the method is carried out without sample analyses that are analyzed outside the process stream.

[0033] It should be noted that the method according to the invention comprises the combination of different analytical methods that are in direct interaction with the process stream. The resulting analyses are based on virtually identical samples of the process stream. This results in synergistic effects regarding the sensitivity of the method according to the invention. Further technical advantages are associated with the feedback provided by the fact that the data can be used to operate the process control.

[0034] In contrast, devices used as offline devices are independent of the process stream line and have no connection to the process stream. Therefore, devices operated as offline devices, i.e., in an ex situ process or outside the process stream, cannot be used to analyze the process stream using an in-situ method.

[0035] In a preferred embodiment, the analysis units, online gas chromatograph (1) and the online IR spectrometer (2), are coupled to a common process control unit (4), wherein the process control unit is operatively connected to a process chamber (11) and monitors, controls or regulates the process taking place in the process chamber.

[0036] Preferably, the method is carried out for the investigation of process streams in which the number of hydrocarbon-containing components in the individual process streams is ≥ 5, more preferably the number of hydrocarbon-containing components in the individual process streams is ≥ 5 to 300. Even more preferably the number of hydrocarbon-containing components in the individual process streams is ≥ 10 to 250.

[0037] In a preferred embodiment, it is a method which further comprises the following steps: c) Evaluating the spectral data obtained during the analytical characterization of the process stream by means of an online IR spectrometer as a function of the time at which this spectroscopic analysis of the process stream was carried out, d) Evaluating the chromatography data obtained during the analytical characterization of the process stream by means of the online gas chromatograph as a function of a sampling time for samples taken from the process stream, e) Training a model based on machine learning which models a mathematical relationship between spectral data and corresponding chromatography data with regard to a same process stream, using the evaluation results obtained in steps c) and d) with regard to the process stream passed through the process stream line in step b).

[0038] It is preferred that the model training takes place in situ—during the reaction or during the process. This has the advantage that external calibration, which can be costly, is unnecessary. The model training therefore takes place in situ during the reaction, and costly external calibration is not necessary. The model can be adapted and supplemented for each new process stream, or a new model can be trained, minimizing the risk of overfitting.

[0039] Such a training dataset is preferably obtained as follows: 1) For rapid changes in the process stream (< 30 min): Conduct the reaction under comparatively "mild" process conditions, which slow the rate of change but still yield representative product compositions (low WHSV, temperature, etc.). If necessary, additional variation of the process parameters is also required to obtain a broader range of product compositions. 2) For slow changes in the process stream (> 30 min): In principle, it would be possible to work exclusively with GC if the reaction proceeds slowly. However, rapid analysis offers opportunities to increase the degree of parallelization, to capture different product streams, or to achieve a target parameter more quickly when controlling for this parameter.

[0040] In embodiments of the method according to the invention, after receiving the training data set in step e) or according to steps c) - e), the chromatographic method used to determine the product composition can be paused, and the measurement is carried out exclusively using the spectroscopic method. (However, gas chromatography can be used for spot checks.)

[0041] During a training phase, representative ranges of compositions of the process stream are collected from the analysis results; these can be supplemented by varying process parameters such as temperature, pressure, WHSV or the reactant composition.

[0042] In case of (too) rapid temporal change of the process stream, which would result in insufficient temporal resolution for gas chromatographic analysis methods, the deactivation can be slowed down by appropriate variation of process parameters (e.g. by variation of WHSV, T, H 2 partial pressure, water vapor addition, etc.).

[0043] A particularly preferred embodiment of the method is one in which, preferably during the training phase, steps c) - e) are carried out and at least one reaction parameter is changed compared to the same reaction parameter as set in steps a) and / or b), wherein this parameter is preferably selected from: WHSV, temperature, total pressure, and / or partial pressure of reactants. By specifically changing the reaction parameters, the reaction rate or conversion can be reduced or increased, respectively.

[0044] These reaction or process parameters are preferably one or more parameters selected from the group consisting of WHSV, T, total pressure, partial pressures of the reactants, and the addition of co-feed molecules, for example, water as a co-feed molecule. By characterizing the process stream obtained for a set of process parameters, a data set of analytical characterization data is obtained that is characteristic of the respective parameter space.

[0045] It is preferred that the process parameters lie in a process parameter space that is of technical importance for the process under investigation.

[0046] Preferably, the temperature of the process coupled with the method according to the invention is in a range of 50 - 1000 °C, more preferably in a range of 150 - 750 °C, further preferably in a range of 250 - 650 °C.

[0047] The process pressure is preferably in a range of 0.5–500 bar, more preferably in a range of 1–250 bar, and even more preferably in a range of 5–200 bar. The pressure specified here refers to the absolute pressure value (dhpa).

[0048] The WHSV is preferably in a range of 0.01 - 500 h -1< , more preferably in a range of 0.5 - 20 h -1< , furthermore preferably in a range of 1 - 5 h -1< .

[0049] In a preferred embodiment, the process according to the invention is characterized in that it comprises at least two different phases, one of which is

[0050] Phase is a training phase comprising steps a) to e) and the second phase is an actual measuring phase in which an analytical characterization of a process stream conducted through the at least one process stream line is carried out by means of the online IR spectrometer (2) on the basis of the model trained in the training phase.

[0051] In a preferred embodiment, the method according to the invention distinguishes between a training phase and a measurement phase. In the measurement phase, all reaction parameters typically lie within a target parameter space. In the training phase, however, the reaction parameters typically do not yet lie within the target parameter space, or at least one reaction parameter is varied in a controlled manner to generate a representative range of different process streams with different compositions.

[0052] In the process according to the invention, it is also preferred that the process stream conducted through the process stream line is a gaseous process stream; the temperature of the gaseous process stream is preferably in the range of 20–350°C, more preferably in the range of 50–220°C. The temperature control of the process stream line ensures that the hydrocarbon-containing components in the process stream are present in the gas phase. This can save the time required for condensation of the hydrocarbon-containing components. In preferred embodiments, the temperature is adjusted to achieve a signal-to-noise ratio in which the coupling of oscillation bands is reduced.

[0053] In the case that the online gas chromatograph (1) and the online IR spectrometer (2) are arranged serially with respect to the process stream line, it is preferred that the measurements are carried out with a time offset, wherein the time offset is in the range of 1 sec to 180 sec.

[0054] A preferred embodiment of the method according to the invention comprises feedback. In a preferred embodiment, the method is characterized in that the analysis units, online gas chromatograph (1) and the online IR spectrometer (2), are coupled to a common process control unit (4), wherein the process control unit (4) is operatively connected to a process chamber (11) and monitors, controls, or regulates the process taking place in the process chamber. The measured data determined are used to control or regulate the process. The short measurement times are particularly advantageous because they enable rapid feedback, which is in the range of the duration of the online IR measurement. The reason for this is that the time required for processing the data and transmitting control or regulation signals to process units is very short, amounting to only milliseconds.

[0055] It is preferred that the process control unit (4) controls the process in such a way that the product structure is controlled by adapting a process operating parameter, preferably the product structure is controlled in such a way that the octane number is constant, that the selectivity is constant or that the conversion is constant, further preferably the control is carried out by changing the parameters from the group temperature and WHSV.

[0056] The method for investigating at least one process stream is preferably used for investigating a process stream which originates from the group selected from methanol conversion processes such as MTO (e.g. methanol-to-olefins), dehydrogenation reactions (e.g. propane dehydrogenations), coupling reactions (e.g. methane couplings), naphtha reforming and processes for the synthesis and conversion of aromatics (e.g. transalkylations, alkylations, dealkylations).

[0057] In the method for investigating at least one process stream, the online IR spectrometer (2) used is preferably a device which operates in the mid-IR range (MIR) and in which the wavenumbers are in the range of 400 cm -1< - 3500 cm -1<.

[0058] In the method according to the invention for investigating at least one process stream, it is also preferred that the model or method used in step e) comprises a statistical method selected from the group of multivariate analyses such as principal component analysis (PCA), partial least squares regression (PLS), principal component regression (PCR), multilinear regression analysis (MLR), discriminant analysis, or neural networks. A correlation between the analytical results obtained by means of online IR spectroscopy and online gas chromatography is achieved by relating the analytical results to one another using suitable statistical methods and training a model.Suitable static methods are, for example, those selected from the group of multivariate analyses such as principal component analysis (PCA), partial least squares regression (PLS), principal component regression (PCR), multilinear regression analysis (MLR), discriminant analysis or neural networks.

[0059] In the method according to the invention for investigating at least one process stream, it is further preferred that additional data for training the model, which are provided according to steps c) to e), are obtained by varying process parameters such as the temperature, the pressure, the partial pressure of the reactants or the WHSV.

[0060] In a preferred embodiment, the method for investigating at least one process stream relates to a process stream originating from a catalytic process in which a solid catalyst is arranged within the process space, wherein the mass of the solid catalyst arranged within the process space is in the range 0.1-200 ccm, preferably between 0.2 - 20 ccm.

[0061] The present invention also relates to the use of the method according to the invention in the different embodiments presented in the present description for high-throughput testing of at least four, preferably at least eight, more preferably at least twelve, catalysts arranged in parallel reactors.

[0062] When operating high-throughput equipment, the use of powerful analytical methods is of interest for qualitatively and quantitatively determining the reaction products. In a preferred embodiment, the method according to the invention is used in high-throughput testing and in conjunction with the investigation of catalysts. It is also preferred that the catalysts be present in parallel process chambers.

[0063] The process spaces are preferably reactors, and catalysts are preferably arranged in the reactors. The method according to the invention is preferably carried out for investigating process streams in which the process stream line (35) is operatively connected to four reactors arranged in parallel. Further preferably, the process stream line (35) is operatively connected to at least eight reactors arranged in parallel, further preferably at least twelve reactors arranged in parallel, and further preferably to sixteen reactors arranged in parallel. The reactors are preferably tubular reactors.

[0064] The combination of the method according to the invention in conjunction with high-throughput research and high-throughput equipment represents a preferred embodiment because it results in synergistic effects, for example, such that data sets recorded during the training phase or the calibration phase can be used more quickly and with improved efficiency. Parallel process chambers, preferably reactors loaded with different catalysts, provide process streams that exhibit differences and can be used for calibration. The parallel implementation allows the calibration phase to be integrated into the measurement process.

[0065] The use of the method according to the invention is particularly advantageous in conjunction with a high-throughput apparatus, since the method has a high degree of flexibility as regards the chemical process and the high-throughput apparatus can be easily converted if it has to be converted from one chemical reaction process to another, which is often the case, particularly in research operations.

[0066] Using the method according to the invention, reaction networks and processes can be investigated with improved data depth and mechanisms of catalyst deactivation can be better understood, which could potentially form the basis for the development of better catalysts and thus lead to savings of energy and resources.

[0067] In a preferred embodiment, the method according to the invention for investigating process streams additionally comprises one or more process steps for regenerating deactivated catalyst.

[0068] Regeneration primarily removes deposits that have accumulated on the catalyst. To remove the deposits, for example, the temperature of the catalyst can be increased or the deposits can be burned off in the presence of an oxygen-containing gas stream. In coke burn-off studies, the analysis can be specifically focused on the product components CO, CO2, and H2O.

[0069] In a preferred embodiment, the method is used to investigate reaction kinetics. It is further preferred to use the method in conjunction with processes that have short contact times or short deactivation times.

[0070] In preferred embodiments, the IR measurement time per measurement point is ≤ 10 sec, ≤ 5 sec, ≤ 2.5 sec, ≤ 1.25 sec. The term IR measurement time refers to the total measurement duration and includes the time for scanning as well as the time for data evaluation.

[0071] In embodiments, particularly for kinetic investigations, the IR measurement time per measuring point is in the range of tenths of a second, whereby initially only scans are recorded, which are then mathematically processed after the recording.

[0072] The duration of the IR measurement time represents a quantity which is characteristic of the time resolution of the method according to the invention with regard to the detection of changes in the composition of the process stream and the associated process.

[0073] The method according to the invention is suitable for investigating long-term or short-term processes. An advantageous aspect of the method according to the invention is that it can be used flexibly in terms of time: The method according to the invention can be used either to control and monitor long-term processes or industrial plants, or it can be used to conduct short-term experiments. Long-term processes preferably range from 2 weeks to 2 years. Short-term processes preferably range from 2 minutes to 300 hours.

[0074] Short-term processes are preferably characterized by the fact that the catalyst is deactivated within 1-100 minutes, more preferably within a time range of 1.5-60 minutes, and furthermore within a time range of 2-30 minutes. This results in the advantageous aspect that the training phase and the measurement phase are combined into a single training and measurement phase. This advantageous aspect of short-term investigations is of technical importance and great benefit, particularly in combination with regeneration processes. The combination with regeneration processes is achieved by the method first comprising a conversion process, which is then followed by a regeneration process. The sequence of conversion process and regeneration process is preferably repeated, with the repetition being a cyclical mode of operation.Preferably, the number of repetitions or cycles is ≥ 2, more preferably the number of cycles is ≥ 5, even more preferably the number of cycles is ≥ 10.

[0075] The training phase and the measurement phase differ in that the process parameters lie within the target parameter space during the measurement phase. During the training phase, it is preferable that at least one process parameter exhibits a significant deviation and is not within the target parameter space.

[0076] In a preferred embodiment, the method according to the invention relates to a method for high-throughput testing of a plurality of catalysts present in parallel process spaces. The parallel process spaces are preferably parallel reactors. In this preferred embodiment, it is conceivable that the deviation in the process parameter is achieved by carrying out the method with parallel reactors, wherein the variation in the process parameter occurs by filling the parallel reactors with different catalyst materials and / or with different amounts of catalyst material. Thus, in this preferred embodiment of the method, the deviation or variation in the process parameter can be generated within the system.The implementation of the procedure and the calibration included therein during the training phase are also illustrated by the experimental example A.2, which relates to the refinement of aromatic mixtures.

[0077] The process described in Example A.2 was carried out with four different catalysts. The conversion process was initially carried out by storing all four different catalysts at the same temperature. The calibration measurements were further enhanced by conducting measurements at different temperatures. Thus, the process parameters "catalyst" and "temperature" were varied to collect sufficient data during the training phase.

[0078] Thus, it is preferable that both GC characterizations and IR characterizations of the process stream be performed during the training phase when individual selected process parameter spaces are available. The duration of the GC method depends, among other things, on the complexity of the process stream composition and the desired objective for resolution with respect to the individual components. Thus, the duration of the training phase is decisively determined by the duration of the GC method.

[0079] Preferably, 4 or more chromatograms are recorded during the training phase, more preferably the number of chromatograms during the training phase is ≥ 8, even more preferably the number of chromatograms during the training phase is ≥ 15.

[0080] Another characteristic of the method according to the invention is the ratio of IR measurements to GC measurements. Preferably, the ratio of IR measurements to GC measurements is ≥ 6, more preferably, the ratio of IR measurements to GC measurements is ≥ 60, even more preferably, the ratio of IR measurements to GC measurements is ≥ 90, and more preferably, the ratio of IR measurements to GC measurements is ≥ 120. Examples A.1 Conversion of methanol to olefins

[0081] To illustrate the process according to the invention, studies were conducted on the catalytic conversion of methanol to olefins. The studies were carried out using a high-throughput catalyst testing apparatus that could be equipped with up to sixteen reactors arranged in parallel.

[0082] The schematic structure of the apparatus is shown inFigure 9 The apparatus shown in the illustration comprises five reactors (11) - (15) arranged in parallel. The outlet lines (21) - (25) of the reactors are connected to a multiport valve (33) leading to a process flow line (35) that is operatively connected to an online IR spectrometer (2) and an online gas chromatograph (1). Both the online IR spectrometer (2) and the online gas chromatograph (1) are connected to the process control of the device. Two different zeolite-containing catalyst samples were used in the testing, which are commercially available and are referred to as Sample C1 and Sample C2 in the description. Figure 9It can be seen that the process control unit (4), which is connected to the online IR (2) and the online GC (1), has an operative connection to each of the individual reaction chambers (11) - (15). Not visible is the operative connection of the process control unit (4) to the reactant supply, which is preferably also connected to the process control unit (4).

[0083] Sample C1 contained a SAPO-34-based catalyst, and sample C2 contained a ZSM-5-based catalyst. To prepare for the catalytic tests, the individual catalyst samples were mixed with quartz powder and introduced into tubular reactors as powder beds. The catalyst sample quantities used were 0.92 g or 1.85 g. Thus, a total of four catalyst-containing sample mixtures were prepared. For comparison purposes, two reactors were loaded with quartz powder as inert material, which was free of catalyst.

[0084] The reactors used had a tube length of 30 cm and an inner diameter of 15 mm.

[0085] A Hewlett Packard HP 5890 chromatograph equipped with a 20-meter fused-silica column and FI detector was used as the online-IR spectrometer (1). A BRU-KER FTIR spectrometer optimized for the MIR spectral range in the wavenumber range of 7000–400 cm -1 was used as the online-IR spectrometer (2). The instrument's optics were controlled by a He-Ne laser. The spectrometer's beam path was equipped with a heatable gas measuring cell with an internal volume of 500 mL and an optical beam length of 75 cm. The measuring cell was maintained at a temperature of 180 °C during the investigations.

[0086] When feeding the methanol-containing reactant stream, a distinction was made between the following two dosing processes: Dosing process 1. continuous (quasi-continuous) dosing of the methanol-containing carrier gas stream during a time period in the range of 3 - 30 min and dosing process 2. pulsed dosing of the methanol-containing carrier gas stream during a time period in the range of 5 - 60 seconds.

[0087] After the feed stream had been added, an oxygen-containing gas stream was fed into the reaction chamber to burn off the resulting coke and regenerate the catalyst.

[0088] The process for converting the methanol-containing carrier gas stream was carried out according to the process according to the invention. First, a training phase was carried out. During this training phase, the reactor chambers were successively exposed to three different reactant gas streams, each of which had a lower reactant content than the reactant gas stream intended for the target parameter space. To adjust the reactant content in the reactant gas stream, the gas load, which is defined by the WHSV (weight hourly space velocities), was varied. The gas loads selected for the training phase were characterized by the following three WHSV values: 0.2 h -1< , 0.3 h -1< , 0.4 h -1< .

[0089] During the different dosing processes carried out during the training phase, both gas chromatographic analyses using an online gas chromatograph (1) and spectroscopic analyses using an online spectrometer (2) were carried out on the individual process streams, which were successively discharged from the reaction chambers via the multiport valve (33) into the process gas line (35).

[0090] The individual analytical results obtained using the different methods—i.e., spectroscopy and gas chromatography—as a function of time and experimental parameters were correlated using mathematical models. For example, the GC analyses were used to quantitatively determine the respective amounts of individual substances in the process stream (e.g., the content of methanol, dimethyl ether, methane, ethane, ethene, etc.), and the amounts of substance groups within the process stream (e.g., the content of aromatics, olefins, etc.).

[0091] For quantification in this case, a separate model based on the PLS (partial least squares) method was used for nine individual substances, with 15 components adequately explaining the variance. The wavenumbers were reduced to the ranges 423–1040 cm–1 and 1244–2704 cm–1, as the range between 1040–1244 cm–1 and 2700–3100 cm–1 in particular reached the absorption limit in the concentration range used, which prevented further data analysis. A finer or more precise adjustment of the optical path length to the concentration range should allow these wavenumber ranges to be integrated into the evaluation model.

[0092] Following the training phase, investigations were carried out during the production phase in the presence of a reactant gas stream with a higher gas load than in the training phase, under conditions that were within the target parameter space.

[0093] In the present case, the investigations during the production phase and measurement phase were characterized by a gas load in which the WHSVs were in the range of 2 - 20 h -1<. The analytical characterization of the process stream(s) was carried out using an online IR spectrometer (2) (in this case optimized for the MIR range). Measurement signals were recorded at 5-second intervals to characterize the process stream. The quantitative evaluation of the band ranges using the model created during the training phase allowed conclusions to be drawn about the concentration of individual substances based on the spectra. Thus, the concentration of the individual substances could be predicted using the trained model based on the IR spectra.

[0094] To monitor the data, analytical characterizations of the process stream were carried out using the gas chromatograph (1) at 20-minute intervals during the measurement phase, parallel to the IR measurements. When conducting pulse experiments or pulse dosing, additional gas chromatographic analyses were omitted altogether. An overview of the experiments conducted is shown in Table 1. Three experiments were conducted during the training phase and five experiments were conducted during the measurement phase.

[0095] For the conversion of methanol to olefins, the gas load during the training phase was selected such that the WHSV values ​​were 0.2 h -1< , 0.3 h -1< , and 0.4 h -1< . During the measurement phase, the WHSV values ​​were 5 h -1< , 8 h -1< , 12 h -1< , 16 h -1< , and 20 h -1< . Thus, the process streams were generated during the training phase and then characterized as a function of time, which are characteristic for the three different parameter spaces and selected process parameters. Whether these three different parameter spaces are sufficient depends, among other things, on the process under consideration and its temporal behavior.

[0096] In connection with the method according to the invention, it is preferred that a number of process parameter spaces is examined during the training phase, which is preferably ≥ 2. Further preferably, the number of process parameter spaces used in the training phase for generating and characterizing process streams is ≥ 3, and furthermore, the number of process parameter spaces is ≥ 4.

[0097] During the measurement phase, the WHSV parameter space preferably lies in the range from > 1 h -1< to 20 h -1< . In contrast, during the training phase, the WHSV parameter space preferably lies in the range from 0.1 h -1< to 1 h -1< . During the measurement phase, the WHSV range used can also be considered the target parameter space.

[0098] In example A.1, the WHSV lies significantly outside the target parameter space during the training phase.

[0099] With respect to the lower limit of the target parameter space (i.e. a WHSV of 5 h -1< ) ​​during the measurement phase), the deviation of the process parameter WHSV during the training phase is only 8% of the value compared to the measurement phase. This results in the requirement that the process parameters, or at least one process parameter, lie outside the target parameter space during the training phase. The term "outside the target parameter space" means that - when comparing the training phase and the measurement phase - the deviation of at least one process parameter is ≥ 10%, preferably the deviation is ≥ 20%, moreover preferably the deviation is ≥ 50%, particularly preferred is a deviation of ≥ 75%.

[0100] With respect to A.1, the lower limit of the WHSV in the target parameter space was given by a WHSV of 5 h -1<. The WHSV during the training phase was 0.4 h -1<. Thus, the deviation was 4.6 h -1<, which corresponds to a percentage deviation of 92%. With respect to the upper limit of the target parameter, the deviation is given by a WHSV difference of 19.6 h -1<, which corresponds to a percentage deviation of 98%. Table 1 provides an overview of the test numbers and the associated process parameters. Experiment number WHSV [h -1< ] Temperature [°C] Operating mode phase C1_01 0,2 400 continuously Training phase C1_02 0,3 400 continuously Training phase C1_03 0,4 400 continuously Training phase C1_04 5 400 continuously Measurement phase C1_05 8 400 continuously Measurement phase C1_06 12 400 continuously Measurement phase C1_07 16 400 continuously Measurement phase C1_08 20 400 continuously Measurement phase Table 2 shows the percentage of variance covered by the model depending on the components (factors). Components ethylene Propen Butene methane Ethan propane butane Methanol DME 1 0,405 7,72 59,059 28,49 89,47 52,08 47,89 22,06 70,82 2 75,92 92,01 79,86 28,91 89,49 89,06 86,22 22,22 91,81 3 94,5 94,05 96,05 40,55 89,95 97,31 96,56 25,23 96,66 4 94,78 97,52 98,9 94,36 89,96 98,69 97,86 36,79 99,22 5 96 98,65 99,24 95,75 90,27 98,86 97,88 94,43 99,23 6 98,94 99,09 99,5 96,08 97,36 99,08 99,03 95,55 99,43 7 99,12 99,11 99,67 98,06 97,38 99,09 99,38 95,58 99,66 8 99,13 99,19 99,72 98,18 97,88 99,53 99,47 95,65 99,66 9 99,13 99,32 99,73 98,42 98,7 99,53 99,47 95,67 99,72 10 99,13 99,34 99,75 98,65 98,79 99,62 99,55 95,67 99,73 11 99,17 99,35 99,75 98,69 98,84 99,69 99,55 95,8 99,73 12 99,17 99,37 99,77 98,89 99,33 99,71 99,63 95,8 99,73 13 99,17 99,38 99,78 98,89 99,34 99,71 99,65 95,8 99,73 14 99,19 99,38 99,78 98,89 99,34 99,71 99,65 95,83 99,73 15 99,22 99,4 99,78 98,89 99,34 99,72 99,7 95,98 99,73 A.2 Processes for the refinement of aromatic mixtures

[0101] To illustrate the process according to the invention, catalytic studies on the upgrading of aromatic mixtures were also conducted. For the comparability of catalysts, it is often interesting to operate them under process conditions selected so that the different catalysts and processes lead to the same target parameters. In many processes, conversion represents this target parameter. In naphtha reforming, the octane number represents this target parameter.

[0102] This target parameter can be achieved, for example, by changing the process parameters such as temperature, pressure, WHSV, etc. Advantages associated with the process according to the invention arise, particularly when operating the process according to the invention in a parallel arrangement in which a large number of catalysts are investigated simultaneously and in parallel. In a preferred embodiment, the process also relates to adjusting the target parameter to a fixed value or, in conjunction with a device with a large number of reactors arranged in parallel, to adjusting the target parameters in the parallel processes. In those cases in which catalyst deactivation occurs, the process also relates to readjusting the process in order to achieve and maintain the target parameter despite the occurrence of deactivation.

[0103] In connection with the process, it is important that the adjustment and readjustment occur faster than the temporal change in the product spectrum and the deactivation of the catalysts. Another advantage of the process according to the invention is that performing the online IR measurements during the measurement phase enables precise and rapid analysis and thus also ensures rapid readjustment.

[0104] In the present example, the process according to the invention was demonstrated in conjunction with a catalytic process for the conversion and upgrading of an aromatic mixture. The aromatic mixture used was a mixture of mononuclear aromatics having different numbers and positions of alkyl substituents. The aim of the process was to produce a product with a high proportion of p-xylene. The process was carried out in such a way that the conversion of aromatics was selected as the target parameter. The investigations were carried out using a catalyst testing apparatus equipped with four reactors arranged in parallel. Each of the four parallel reactors was filled with a different catalyst material, which are referred to below as catalysts K1 - K4.

[0105] In a first process variant, the process stream was analyzed exclusively using an online gas chromatograph (1), with the spectra being evaluated automatically and the conversion being determined. The duration of each gas chromatographic analysis was approximately 30 minutes. To characterize the process streams of the four parallel reactors sequentially using the single online gas chromatograph (1), a time period of 2 hours was required to obtain a measured value per process stream.

[0106] The four parallel reactors with catalysts K1 to K4 were initially started isothermally at the same temperature, obtaining initial information about the relationship between conversion and temperature. These measurements, which were conducted at the same temperature, were supplemented by additional measurements at different temperatures to obtain a calibration function that showed the relationship between conversion and temperature. The results of the study are presented in the Figure 7shown, with the lower part showing the temperature as a function of TOS (TOS = Time On Stream) and the upper part showing the conversion as a function of TOS. From a TOS of 6 hours, the adjustment to a target conversion value begins using this calibration function, with the result from the gas chromatography analysis being automatically transmitted to the process control and the temperature being adjusted. It can be seen that after about 3 cycles of 2 hours each, the same conversion is achieved and the actual measurement of the catalysts can take place at the same conversion. In the present case, the adjustment process took about 6 hours. Carrying out the method with a device that has a higher degree of parallelization and is equipped with 16 reactors, for example, would then require 16 x 30 min measuring time per cycle, which would result in a duration of 24 hours for three cycles.

[0107] Investigations into the implementation of the procedure in a second variant are described below. Figure 8The process according to the invention is shown with the inclusion of an online IR spectrometer (2). The initial phase, in which the reactors were operated isothermally, was used to record the calibration function. The online gas chromatograph (1) and the online IR spectrometer (2) were operated in parallel. GC chromatograms with the information on the conversion and the corresponding IR spectra were obtained. Corresponding IR spectra means that these were recorded simultaneously with the GC chromatograms. Based on the measurement data obtained, which are given by the GC chromatograms and IR spectra as a function of time, a model could be trained using PLS. Subsequently, the recording of GC chromatograms could be dispensed with when adjusting for constant conversion. The process streams were investigated on the basis of the recording of online IR spectra, determining the conversion using the previously created model.In this example, the measurement time could be reduced from 30 minutes to 2 minutes, and the target conversion was achieved within 24 minutes. In this example, the process according to the invention demonstrated a significant speed advantage over a process carried out without the combination with an online IR spectrometer. To obtain measurement points with higher information density, the online gas chromatograph (1) was switched on when a constant target conversion was reached. The additional use of the online gas chromatograph (1) to characterize the process streams had the advantage of further improving the determination of selectivities. Online IR spectroscopy provided the control parameter, with the conversion being used as the control parameter in this case. An improved determination of selectivities made it possible to improve the differentiation of the catalysts.

[0108] The process for upgrading aromatic mixtures described here as an example can be used in the same way as a process for transalkylation reactions, for the dealkylation of ethylbenzene, for the disproportionation of toluene, for the isomerization of xylenes and in other processes.

[0109] Short description of the characters: Figure 1 shows a series of MIR spectra recorded during the training phase of the process at a WHSV of 0.4 h -1 . This process step is identified in Table 1 with the number Experiment C1_03. The time intervals for recording the spectra were 5 seconds, with the spectra shown being those recorded over a period of several minutes. Figure 2 shows the composition of the process stream as a function of time, which was determined during the training phase by GC at three different gas loads and which was calculated for MIR spectra. The WHSVs were 0.2 h -1 , 0.3 h -1 , and 0.4 h -1 , and were recorded in the experiments with the numbers C1_01, C1_02, and C1_03. Figure 2 upper panel: yields of gas chromatographic analysis; Figure 2Lower panel: predicted yields of the MIR spectra. Figure 3 shows data obtained during the training phase at low WHSVs, in an overlaid representation of GC and MIR data. The large symbols characterize the yields determined by gas chromatographic analyses; the small symbols characterize the predicted yields determined from the MIR spectra. Figure 4 shows the models obtained by relating GC data and IR data during the training phase. The predicted yields are plotted against the measured yields for nine different individual substances. Figure 5 shows the RMSEP (root mean square error predicted) of the cross-validation for the trained models of the nine different individual substances plotted against the number of components, with up to 15 components used.Figure 6 shows the yields predicted from the MIR spectra during the production phase for WHSVs that were within the target parameter space (i.e., high WHSVs were at 5, 8, 12, 16, and 20 h -1 < in five experiments). The large symbols represent yields from gas chromatography analysis; the small symbols represent predicted yields from the MIR. Figure 7 shows values ​​for conversion and temperature as a function of time-on-stream (TOS) obtained during a process for the conversion of aromatics in which four catalysts were arranged in a four-fold parallel reactor system. Conversion is shown at the top; reactor temperature is shown at the bottom. Conversion was determined exclusively using gas chromatography. Figure 8 shows values ​​for conversion and temperature as a function of time-on-stream (TOS) obtained during the process according to the invention, in which four catalysts were investigated in four parallel reactors.Conversion is shown above, reactor temperature is shown below. Conversion was determined using an online gas chromatograph (1), and adjustment to constant conversion was carried out using an online IR spectrometer (2). Figure 9 shows a schematic representation of the device according to the invention in an embodiment equipped with five reaction chambers (11) - (15) arranged in parallel, wherein the online spectrometer (2) and the online gas chromatograph (1) are arranged serially in the process stream line (35), and the process control unit (4) is coupled to the reaction chambers. Figure 10 shows a schematic representation of the device according to the invention in an embodiment which corresponds to that in . Figure 9 shown embodiment, wherein the process stream line (35) is divided into two lines and the online IR (2) and the online GC (1) are arranged in parallel in these two lines. List of reference symbols 1 - online GC, online gas chromatograph 2 - online-IR, online-IR spectrometer or online spectrometer, preferably optimized for the MIR range 4 - Process control unit 35 - Process power line 34 - Exhaust air duct 33 - Multiport valve 21, 22, - 25 - Reaction chamber outlet-side process gas lines in conjunction with multiport valve 11, 12, - 15 - Process rooms, preferably reaction rooms, further preferably tubular reactors

Claims

1. A method for investigating at least one process stream deriving from a catalytic process and comprising at least five different hydrocarbon-containing components, wherein the method comprises at least the steps of: a) providing at least one process stream conduit (35) which is in operative connection with at least one online IR spectrometer (2) and in operative connection with at least one online gas chromatograph (1); b) passing the at least one process stream through the at least one process stream conduit (35), wherein during this passing of the process stream through the process stream conduit (35) an analytical characterization of the process stream using an online IR spectrometer (2) and an online gas chromatograph (1) is performed; c) evaluating the spectral data obtained in the analytical characterization of the process stream using an online IR spectrometer (2) as a function of the time at which this spectroscopic analysis of the process stream was carried out, d) evaluating the chromatography data obtained during the analytical characterization of the process stream using the online gas chromatograph (1) as a function of a sampling time for samples taken from the process stream, e) machine learning-based training of a model that models a mathematical relationship between spectral data and corresponding chromatography data in respect of an identical process stream by using the evaluation results obtained in steps c) and d) in respect of the process stream passed through the process stream conduit in step b) .

2. The method for investigating at least one process stream derived from a catalytic process according to claim 1, wherein the method comprises a feedback and wherein the online gas chromatograph (1) and the online IR spectrometer (2) are coupled to a common process control unit (4), wherein the common process control unit (4) is in operative connection with a process space (11) and controls, governs or regulates the process proceeding in the process space.

3. The method according to claim 1 or claim 2, wherein steps c) - e) are performed and at least one reaction parameter is altered in relation to the same reaction parameter as set in steps a) and / or b), wherein this parameter is selected from: WHSV, temperature, total pressure and / or partial pressure of reactants.

4. The method for investigating at least one process stream according to claim 1 to claim 3, wherein it comprises at least two different phases, wherein one phase is a training phase comprising steps a) to e) and the second phase is an actual measurement phase in which an analytical characterization of a process stream passed through the at least one process stream conduit is carried out using the online IR spectrometer (2) on the basis of the model trained in the training phase.

5. The method for investigating at least one process stream according to at least one of the preceding claims, wherein the process stream passed through the process stream conduit (35) is a gaseous process stream, the temperature of the gaseous process stream preferably being in the range of 20 - 350°C.

6. The method for investigating at least one process stream according to any of claims 1-5, wherein the online gas chromatograph (1) and the online IR spectrometer (2) are serially arranged in respect of the process stream conduit and the temporal offset is in the range of 1 sec to 180 sec.

7. The method for investigating at least one process stream according to any of claims 1-6, wherein the analysis units, the online gas chromatograph (1) and the online IR spectrometer (2), are coupled to a common process control unit (4), wherein the process control unit is in operative connection with a process space (11) and controls, governs or regulates the process proceeding in the process space.

8. The method for investigating at least one process stream according to claim 7, wherein the process control unit (4) regulates the process such that the product structure is controlled by adapting a process operating parameter.

9. The method for investigating at least one process stream according to any of claims 1 to 8, wherein the process comprises at least one process from the group selected from methanol conversion processes such as MTO (methanol to olefins), dehydrogenation reactions such as propane dehydrogenation, coupling reactions such as methane coupling, naphtha reforming and processes for conversion of aromatics.

10. The method for investigating at least one process stream according to any of claims 1 - 9, wherein the online IR spectrometer (2) operates in the mid IR range (MIR) and the wavenumbers are in the range of 400 cm-1 - 3500 cm-1.

11. The method for investigating at least one process stream according to any of claims 1 - 10, wherein the model or the method as employed in step e) comprises a statistical method selected from the group of multivariate analyses such as principal component analysis (PCA), partial least squares (PLS) regression, principal component regression (PCR), multi-linear regression (MLR) analysis, discriminant analysis or neural networks.

12. The method for investigating at least one process stream according to any of claims 1 - 11, wherein additional data for training the model, provided for according to steps c) and e), are obtained by variation of process parameters such as temperature, pressure, partial pressure of the reactants or WHSV.

13. The method for investigating at least one process stream according to any of claims 1 - 12, wherein the mass of the catalyst employed for the process is in the range of 0.1-200 ccm.

14. The use of the method according to any of the preceding claims 1 to 13 in high-throughput testing of at least four, preferably at least eight, more preferably at least 12, catalysts arranged in parallel reactors.