A method for detecting the formation of a deposition layer on a cooled surface inside an agitated vessel, such as in a scraped heat exchanger

Abstract

The present invention relates to a method for detecting the formation of a deposition layer on a cooled surface being provided inside a vessel comprising at least one agitator, in particular for detecting the formation of a crystal layer on a cooled surface being provided inside a scraped heat exchanger and preferably in a scraped crystallizer, wherein the method comprises the following steps: a) operating the vessel being filled with a liquid, a suspension or an emulsion by driving at least one agitator and by cooling at least one surface and b) time-resolved determining during the operation of the vessel in step a) the acceleration of the vessel and determining, whether the acceleration of the vessel has been increased within a first predetermined time period by more than a predetermined threshold, wherein the length of the first predetermined time period is at most 1 hour and the predetermined threshold is an increase of the acceleration of the vessel of at least 10% of the acceleration of the vessel before the start of the first predetermined time period.

Description

[0001] Sulzer Management AG S13789P

[0002] A method for detecting the formation of a deposition layer on a cooled surface inside an agitated vessel, such as in a scraped heat exchanger

[0003] The present invention relates to a method for detecting the formation of a deposition layer on a cooled surface being provided inside a vessel comprising at least one agitator, in particular for detecting the formation of a crystal layer on a cooled surface being provided inside a scraped heat exchanger and preferably in a scraped crystallizer. Furthermore, the present invention relates to a respective vessel.

[0004] Vessels comprising one or more cooled surfaces and at least one agitator are frequently used in industrial processes, such as for instance as crystallizer, as heat exchanger or the like. Specific examples therefore are scraped heat exchangers and scraped crystallization apparatuses, such as scraped suspension crystallization apparatuses, in which knives are rotated over the cooled surfaces in order to avoid the formation of a deposition layer, such as a crystal layer, on the cooled surfaces. In contrast to layer crystallization processes, the formation of crystal layers on the cooled surfaces of the crystallizer is not intended during a suspension crystallization. During the suspension crystallization a melt or solution, which includes the target compound and at least one different compound as impurity, is cooled in a vessel so that crystals are formed within the melt or solution so as to thereby form a suspension of crystalline particles, in which the target compound is enriched, being suspended in the melt or solution being depleted of the target compound. After completion of the crystallization, the crystals are separated from the melt or solution, e.g. by filters, centrifuges or other equipment for sol- id / liquid separation, and, if necessary, are further purified, such as in a second crystallization stage or by means of another suitable purification method. The required cooling of the suspension is achieved during the suspension crystallization by cooling one or more surfaces being provided within the vessel, such as the inner wall of the vessel. More specifically, a cooling agent is circulated on the side of the cooled surface being opposite to the side being in contact with the suspension.

[0005] The formation of a crystal layer on the cooled surface, which is also denotated as fouling, as layering or as layer formation, respectively, severely affects the suspension crystallization process, since the crystal layer being formed on the cooled surface significantly reduces the heat transfer through the cooled surface and thus the cooling of the suspension within the vessel. However, the driving force for the formation of a crystal layer on the cooled surface is quite high. In order to avoid a formation of a crystal layer on the cooled surface during the suspension crystallization process, the crystallization is usually performed in a scraped surface crystallizer, in which knives are continuously rotated over the cooled surfaces to thereby scrape off any deposits, such as crystal layers. Thus, the rotating knives are intended to keep the heat exchange resistance of the cooled surface low and to remove crystals from the cooled surface, which can then grow in the suspension so as to further increase the product mass. In addition to reducing the heat transfer through the cooled surface, crystal layers being formed on the cooled surfaces further create a significant increase of the resistance for the rotation of the knives, thereby avoiding that the knives may fulfil their function. However, the formation of undesired crystal layers on the cooled surface cannot be reliably avoided in particular in cases, in which large driving forces for crystal layer formation as consequence of a desired large productivity and capacity are present. If certain unfavorable operation conditions are applied, such as the adjustment of a too high cooling, large concentrations of the target component or a too low knife rotational speed, wherein as a consequence thereof the driving force for the formation of a crystal layer on the cooled surface is increased, a crystal layer formation on the cooled surfaces starts and accelerates during the operational time so that after a certain time a thick solid crystal layer is formed on the cooled surface, which cannot be removed anymore by the rotating knives. The heat transport resistance increases continuously with the increment of the thickness of the crystal layer until an equilibrium between heat input and output is reached and the process productivity is reduced to virtually zero. The crystallization process must then be stopped and the crystal layer must be removed, e.g. by increasing the operation temperature above the melt temperature of the pure target component.

[0006] On account of these reasons, it is desirable to monitor the possible crystal layer formation on cooled surfaces within the vessel during the suspension crystallization process. However, a direct measurement or detection of the formation of a crystal layer on the cooled surfaces of a scraped crystallizer is not possible. Hence, indirect process information are usually used to detect the formation of a crystal layer, such as the power consumption of the scraper motor and / or the temperature evolution of the suspension. If the temperature does not decrease anymore during the cooling and the power consumption of the scraper motor increases, an undesired layer formation is one possible cause. However, only using one or two of the aforementioned indicators, i.e. either the power consumption of the scraper motor and / or the temperature evolution of the suspension, is usually not sufficient to detect a crystal layer formation with a reasonable reliability. Additionally, a certain crystal layer thickness must be present in order to lead to a change of the motor power input or temperature evolution of the suspension so that with these techniques a crystal layer formation on a cooled surface within the vessel can be only detected, if at all, quite late, when the crystal layer has already a considerable thickness.

[0007] In view of this, the object underlying the present invention is to provide a simple and cost-efficient method, which allows to precisely and reliably detect the formation of a deposition layer on a cooled surface being provided inside a vessel comprising at least one agitator, in particular for detecting the formation of a crys- tai layer on a cooled surface being provided inside a scraped heat exchanger and preferably in a scraped crystallizer already at an early stage, i.e. when the deposition layer, such as crystal layer, is still comparable thin and does not extend over the whole of the cooled surface.

[0008] In accordance with the present invention this object is satisfied by providing a method for detecting the formation of a deposition layer on a cooled surface being provided inside a vessel comprising at least one agitator, in particular for detecting the formation of a crystal layer on a cooled surface being provided inside a scraped heat exchanger and preferably in a scraped crystallizer, wherein the method comprises the following steps: a) operating the vessel being filled with a liquid, a suspension or an emulsion by driving at least one agitator and by cooling at least one surface and b) time-resolved determining during the operation of the vessel in step a) the acceleration of the vessel and determining, whether the acceleration of the vessel has been increased within a first predetermined time period by more than a predetermined threshold, wherein the length of the first predetermined time period is at most 1 hour and the predetermined threshold is an increase of the acceleration of the vessel of at least 10% of the acceleration of the vessel before the start of the first predetermined time period.

[0009] This solution bases on the finding that the monitoring of the acceleration of the vessel allows to reliably detect the formation of a deposition layer, such as crystal layer, on a cooled surface being provided inside an agitated vessel already at a very early stage, i.e. when the deposition layer is still comparable thin and does not extend over the whole of the cooled surface. Any agitated system has a certain eigenfrequency due to the moving agitator. For instance, a suspension crystallization apparatus comprising a scraped suspension crystallizer being provided with a motor for rotating the scraper knife and being further connected with a wash column comprising a piston, which reciprocatingly moves up and down the wash column, and being further provided with a cooling unit, a circulation pump, a fan and a piping system has an eigenfrequency, which is influenced by the vibrations resulting from the rotating knives at least temporarily contacting the cooled surface, by the vibrations resulting from the piston movement in the wash column, by the vibrations of the motor driving the scraper knives, by the vibrations of the cooling unit, by the vibrations of the piping system and by the vibrations of the circulation pump. It has been surprisingly found by the inventors of the present invention that the acceleration of an agitated vessel drastically increases within a very short time interval, when crystals start to deposit on a comparable low portion of the cooled surface, such as that about 10% of the cooled surface is covered by a thin deposition layer or crystal layer, respectively. On account of this reason, the detection of an acceleration increment allows to reliably detect the formation of a crystal layer on the cooled surface of a vessel at a very early stage. A further advantage of the method in accordance with the present invention is, that it is simple and cost-efficient, because the monitoring of the acceleration of the vessel can be easily and cost-efficiently determined by using one or more accelerometers, which are also denotated as vibration sensors.

[0010] The term agitator as used herein means in accordance with the present invention any moved means, such as a rotating knife, an impeller agitator or the like.

[0011] Furthermore, the terms deposition layer and crystal layer are broadly defined within the present invention as any deposit or crystal being formed on or in close vicinity of a cooled surface within the vessel and having a more than molecular size, such as having a size so as to cover at least 0.01 %, such as for instance 1 %, 5% or 10% of the cooled surface. Hence, a single deposit or a single crystal having such a size and being arranged on the cooled surface is according to the present invention a deposition layer or crystal layer, respectively. As set out above, the method comprises a step a) of operating the vessel and a step b) of time-resolved determining or measuring, respectively, the acceleration of the vessel during the operation of the vessel in step a). However, steps a) and b) are not subsequent steps, but are performed in parallel so that the denotations step a) and step b) are only used in order to easily refer to one of these actions.

[0012] In accordance with the present invention the acceleration of the vessel is time- resolved determined during the operation of the agitated vessel and it is further determined, whether the acceleration of the vessel has been increased within a first predetermined time period of at most 1 hour by more than a predetermined threshold of at least 10% of the acceleration of the vessel before the start of the first predetermined time period. Determining, whether the acceleration of the vessel has been increased within the first predetermined time period of at most 1 hour by more than the predetermined threshold of the acceleration of the vessel before the start of the first predetermined time period means that the acceleration of the vessel being measured during the first predetermined time period is compared with the acceleration of the vessel having been measured before the start of the first predetermined time period, i.e. within a reference time period or reference time interval, respectively, and that it is then determined, whether the difference between the two acceleration numeric values is at least the predetermined threshold or less. Reference time interval means herein any time interval, which ended before the start of the first predetermined time period. If the acceleration of the vessel has been increased within the first predetermined time period by more than the predetermined threshold, it is decided that one or more deposition layers, in particular crystal layers, have been formed on one or more of the cooled surfaces being provided inside the vessel. Then, a suitable measure may be taken, such as stopping the operation of the vessel immediately or within a certain time period and removing the crystal layer, before restarting the process. Since the acceleration of an agitated vessel usually slightly changes periodically or irregularly even at very short time intervals, the measured acceleration is - even if measured within short subsequent time intervals - not constant. Therefore it is preferred that, if the acceleration is not constant within the length of the time interval, such as during the first predetermined time period, the average acceleration during the measurement time interval is used, which is calculated by summing up all single numeric values for the acceleration of the vessel measured during the length of the time interval, such as the first predetermined time period, and then by dividing the obtained sum by the number of single numeric values measured for the acceleration of the vessel during the time interval.

[0013] In accordance with the present invention it is determined in step b) whether the acceleration of the vessel has been increased within the first predetermined time period of at most 10 second by more than the predetermined threshold of at least 10% of the acceleration of the vessel before the start of the first predetermined time period. The acceleration of the vessel before the start of the first predetermined time period may be that acceleration being measured directly before the start of the first predetermined time period. In this embodiment, it is preferred that it is determined whether the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel of a time period, which has ended exactly at the start of the first predetermined time period. Alternatively, it may be determined whether the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel of a time period, which has ended one or more seconds or one or more minutes or even one or more hours or one or more days before the start of the first predetermined time period. Hence, in the latter case the actually measured acceleration of the vessel is compared with a reference numeric value for the acceleration of the vessel, such as a reference numeric value for the acceleration of the vessel, which has been recorded at the beginning of the operation, i.e. within a reference time interval, in order to deter- mine whether the acceleration of the vessel has been increased by more than 10% during the first predetermined time period. However, it is preferred to determine whether the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel of a time period, which has ended exactly at the start of the first predetermined time period, less than one second before the start of the first predetermined time period or less than one minute before the start of the first predetermined time period or less than one hour before the start of the first predetermined time period.

[0014] In accordance with a further preferred embodiment of the present invention, the determination, whether the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period is made by comparing the acceleration or average acceleration, respectively, of the vessel measured within the first predetermined time period with the acceleration or average acceleration, respectively, of the vessel measured within a time period having the same length as the first predetermined time period and having been ended less than 1 hour, preferably less than 10 minutes, more preferably less than 1 minute, still more preferably less than 1 second and most preferably directly at the start of the first predetermined time period.

[0015] Since the acceleration of the vessel surprisingly significantly increases within a very short time after a deposition layer, such as crystal layer, reaches a certain critical extent of for instance about 10% of the cooled surface, it is suggested in a further embodiment of the present invention that the length of the first predetermined time period in step b) is at most 10 minutes, more preferably at most 1 minute, still more preferably at most 30 seconds, yet more preferably at most 10 seconds, yet more preferably at most 1 second, yet more preferably at most 500 milliseconds and most preferably at most 100 milliseconds. The lower boundary for the length of the first predetermined time period in step b) is preferably more at least 0.1 milliseconds, more preferably at least 1 millisecond, still more preferably at least 10 milliseconds and most preferably at least 100 milliseconds.

[0016] A particular reliably determination of the formation of a deposition layer, in particular crystal layer, on the one or more cooled surfaces of the vessel is achieved, when the predetermined threshold for the increase of the acceleration of the vessel in step b) is at least 20%, more preferably at least 50%, still more preferably at least 100% and most preferably at least 200% of the acceleration of the vessel before the start of the first predetermined time period.

[0017] In order to even increase the reliability of the detection of the formation of a deposition layer on the one or more cooled surfaces, it is proposed in a further development of the idea of the present invention that is further determined in step b), whether the increased acceleration measured within the first predetermined time period decreases significantly shortly after the first predetermined time period or not. If the increased acceleration measured within the first predetermined time period does not significantly decrease shortly after the first predetermined time period, it can be reliably excluded that the increase of the acceleration of the vessel during the first predetermined time period was only due to a measurement failure, an artefact or the like. Therefore, it is preferred in this embodiment that it is further determined in step b), whether the acceleration of the vessel has not been decreased within a second predetermined time period by at least 5% of the acceleration of the vessel measured within the first predetermined time period, wherein the second predetermined time period starts not later than 1 hour, preferably not later than 10 minutes, more preferably not later than 1 minute, still more preferably not later than 1 second and most preferably directly at the end of the first predetermined time period. Moreover, it is preferred that the length of the second predetermined time period is at least 100 milliseconds. Again, if the acceleration is not constant during the second predetermined time period, the average acceleration of the vessel is preferably determined in the second predetermined time period, wherein the average acceleration of the vessel is calculated by summing up all single numeric values for the acceleration of the vessel measured during the second predetermined time period and by then dividing the obtained sum by the number of single numeric values measured for the acceleration of the vessel. If the acceleration of the vessel has not been decreased within the second predetermined time period by at least 5% of the acceleration of the vessel measured within the first predetermined time period, it is decided that one or more deposition layers, in particular crystal layers, have been formed on one or more of the cooled surfaces being provided inside the vessel.

[0018] Good results are in particular obtained, when the length of the second predetermined time period in step b) is at least 250 milliseconds, more preferably at least 500 milliseconds, still more preferably at least 1 second, yet more preferably at least 2.5 seconds, yet more preferably at least 5 seconds, yet more preferably at least 10 seconds and most preferably at least 20 seconds. The upper boundary for the length of the second predetermined time period in step b) is preferably at most 1 hour, more preferably at most 30 minutes, still more preferably at most 10 minutes, yet more preferably at most 5 minutes and most preferably at most 1 minute.

[0019] In addition, it is preferred that it is determined in step b), whether the acceleration of the vessel has not been decreased within the second predetermined time period by at least 20%, preferably at least 50% and most preferably at least 90% of the acceleration of the vessel within the first predetermined time period.

[0020] In accordance with another preferred embodiment of the present invention, the (average) acceleration of the vessel is determined in step b) in subsequent time intervals, wherein the numeric values for the (average) acceleration of the vessel measured during the time intervals are at least temporarily stored, wherein each of the time intervals has a length of at most the first predetermined time period, pref- erably of at most 50% of the first predetermined time period, more preferably of at most 25% of the first predetermined time period and most preferably of at most 10% of the first predetermined time period. Thereby, a good time resolved monitoring of the acceleration of the vessel is achieved. Good results are in particular obtained, when the time-resolved determination of the acceleration of the vessel in step b) is performed in measurement time intervals having a length of 10 milliseconds to 10 minutes, more preferably of 100 milliseconds to 1 minute, still more preferably of 1 to 20 seconds and most preferably of 2 to 10 seconds, such as of about 5 seconds. If the acceleration of the vessel is determined in step b) in subsequent time intervals being shorter than the first predetermined time period, the acceleration of the vessel within the first predetermined time period is preferably the average of the accelerations measured during the time intervals forming the first predetermined time period. Likewise thereto, if the acceleration of the vessel is determined in step b) in subsequent time intervals being shorter than the second predetermined time period, the acceleration of the vessel within the second predetermined time period is preferably the average of the accelerations measured during the time intervals forming the second predetermined time period.

[0021] The acceleration signal being obtained during the time-resolved determination of the acceleration of the vessel during step b) in a measurement time interval can be interpreted as the sum of all vibrations each having a certain frequency occurring in the vessel during the respective measurement time interval, i.e. the sum of dozens, hundreds or even thousands different vibrations each having a different frequency. Therefore, the acceleration increase during the first predetermined time period does not need to be a steep rise occurring within one millisecond, but may be an essentially continuous increase over for instance 1 second. In order to receive a more significant acceleration increase, it is suggested in a further development of the idea of the present invention that the acceleration values obtained during the time-resolved determination of the acceleration of the vessel in step b) are converted from the time domain into the frequency domain, before it is deter- mined whether the acceleration of the vessel at a specific frequency or in a specific frequency range has been increased within the first predetermined time period by more than the predetermined threshold of at least 10%. It is preferred that the frequency domain comprises frequencies in a range of 1 Hz to 30 kHz. This embodiment allows a particular precise and reliable determination, whether a deposition layer has been formed on a cooled surface being inside the vessel or not. This is due to the fact that the acceleration increase is highest at frequencies corresponding to or being at least close to the eigenfrequency of the vessel so that for these frequencies a steep rise of the acceleration of the vessel results. Preferably, the conversion of the acceleration values obtained during the time-resolved determination of the acceleration of the vessel in step b) from the time domain into the frequency domain is performed with fast Fourier transformation (FFT), which is an efficient algorithm for calculating the discrete Fourier transform (DFT), which breaks down a discrete-time signal into its frequency components. Preferably, a power spectral density (PSD) is computed from the DFT with the advantage that the amplitude is not influenced by the duration for which the signal is captured. For instance, the numeric values obtained for the acceleration of the vessel for each subsequent 1 -second time interval are converted from the time domain into the frequency domain. For example, the time-resolved determination in step b) is performed with a sampling rate of 1 kHz, which is equivalent to measurement time intervals of 1 millisecond. The acceleration values obtained in 1 ,000 subsequent measurement time intervals are converted with FFT or preferably with a PSD from the time domain into the frequency domain so as to obtain for a first 1 -second time interval the frequency dependent acceleration values of the vessel, which may be visualized as diagram showing the acceleration of the vessel on the ordinate in dependency of the frequency on the abscissa. This is done for the acceleration values obtained in each following 1 ,000 subsequent measurement time intervals so as to obtain the frequency dependent acceleration values of the vessel for subsequent 1 -second time intervals. The obtained frequency dependent acceleration values of the vessel for each 1 -second time interval are compared with the respective frequency dependent acceleration values of the vessel obtained in a previous 1 -second time interval, which is subsequently denoted as 1 -second reference time interval. It is then determined, whether the acceleration at at least one frequency value of the 1 -second time interval has been increased by more than 10% compared to the acceleration at the same frequency of a 1 -second reference time interval and, if so, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period. Alternatively, instead of comparing the single frequencies of the subsequent 1- second time intervals, frequency bands of the subsequent 1 -second time intervals may be compared with each other and thereby it is determined, whether the acceleration at at least one frequency band of the 1 -second time interval has been increased by more than 10% compared to the acceleration at the same frequency band of the 1 -second reference time interval and, if so, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period. Preferably, the reference time interval is chosen at a time, at which for sure no deposition layer has been formed. More specifically, it is preferred that the average acceleration values of each frequency band are calculated at a time interval by summing up all acceleration values of the frequency band of the time interval and by then dividing the sum by the number of frequencies of the frequency band, wherein the average acceleration values of each frequency band of the time interval of a measurement are compared with the respective average acceleration values of the same frequency band of the measurement of a reference time interval. Again, it is preferred that the reference time interval is chosen at a time, at which for sure no deposition layer has been formed. If the average acceleration value of at least one frequency band of the 1 -second time interval has been increased by more than 10% compared to the average acceleration at the same frequency band of the 1 -second reference time interval, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period. Still alternatively, instead of comparing the average acceleration values of the frequency bands, only the highest acceleration value of a frequency band may be compared with the highest acceleration value of the same frequency band of a previous measurement. If the highest acceleration value of at least one frequency band of a 1 -second time interval has been increased by more than 10% compared to the highest acceleration at the same frequency band of a 1 -second reference time interval, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period. Still alternatively, instead of comparing the average acceleration values of the frequency bands, the median acceleration value of a frequency band may be compared with the median acceleration value of the same frequency band of a previous measurement. If the median acceleration value of at least one frequency band of a 1 -second time interval has been increased by more than 10% compared to the median acceleration at the same frequency band of a 1 -second reference time interval, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period. For instance, the frequency bands may lie in a range of 1 Hz to 30 kHz and preferably the difference between the highest frequency and the lowest frequency of each of the frequency bands is 5 to 10,000 Hz, more preferably 10 to 1 ,000 Hz, more preferably 50 to 500 Hz and most preferably 100 to 300 Hz, such as about 200 Hz. Furthermore, it is preferred that the gap between adjacent frequency bands is less than 100 Hz, more preferably less than 10 Hz, still more preferably at most 1 Hz and most preferably 0 Hz. However, it is also preferred that the different frequency bands do not overlap with each other. Irrespective of whether comparing in the aforementioned embodiments acceleration values at specific frequencies or the (average) acceleration values of frequency bands, instead of using subsequent 1 -second time intervals, each time interval may have a length up to the length of the first predetermined time period, i.e. up to 1 hour. Preferably, the respective time intervals each have a length of at most 10 minutes, more preferably of at most 1 minute, still more preferably of at most 30 seconds, yet more preferably of at most 10 seconds, yet more preferably of at most 1 second, yet more preferably of at most 500 milliseconds and most preferably of at most 100 milliseconds, whereas the lower boundary for the length of the respective time interval is preferably at least 0.1 milliseconds, more preferably at least 1 millisecond, still more preferably at least 10 milliseconds and most preferably at least 100 milliseconds. Hence, preferably the respective time intervals have a length of 0.1 milliseconds to 1 hour, more preferably of 1 millisecond to 1 hour, still more preferably of 10 milliseconds to 1 hour, yet more preferably of 100 milliseconds to 1 hour, such as 1 second to 1 minute, or 2 to 20 seconds, or 5 to 10 seconds. Accordingly, it is preferred that in step b) the time-resolved determination of the acceleration of the vessel is measured at measurement time intervals having a length between 10 milliseconds and 1 hour, preferably between 100 milliseconds and 10 minutes and more preferably between 1 and 10 seconds, wherein subsequent measurement time intervals summing up to time intervals of 10 milliseconds to 1 hour and preferably of 1 second to 1 minute are converted with fast Fourier transformation from the time domain into the frequency domain so as to obtain for subsequent time intervals the frequency dependent acceleration values of the vessel.

[0022] Accordingly, it is preferred in the aforementioned embodiments that in step b): i) the obtained frequency dependent acceleration values of the vessel for each time interval are compared with the respective frequency dependent acceleration values of the vessel obtained in a reference time interval and it is determined, whether the acceleration at at least one frequency value of the time interval has been increased by more than 10% compared to the acceleration at the same frequency of the reference time interval and, if so, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period, or ii) frequency bands of the subsequent time intervals are compared with each other and it is thereby determined, whether the acceleration at at least one frequency band of a time interval has been increased by more than 10% compared to the acceleration at the same frequency band of a reference time interval and, if so, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period, or iii) average acceleration values of each frequency band are calculated by summing up all acceleration values of the frequency band and by then dividing the sum by the number of frequencies of the frequency band, wherein the average acceleration values of each frequency band of a time interval are compared with the respective average acceleration values of the same frequency band of a reference time interval and, if the average acceleration value of at least one frequency band of a time interval has been increased by more than 10% compared to the average acceleration at the same frequency band of the reference time interval, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period, or iv) maximal acceleration values of each frequency band are calculated by determining the maximal acceleration values within a frequency band, wherein the highest acceleration value of a frequency band of a time interval is compared with the highest acceleration value of the same frequency band of a reference time interval and, if the highest acceleration value of at least one frequency band of the time interval has been increased by more than 10% compared to the highest acceleration at the same frequency band of the reference time interval, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period, or v) median acceleration values of each frequency band are calculated by determining the median acceleration values within a frequency band, wherein the median acceleration value of a frequency band of a time interval is compared with the median acceleration value of the same frequency band of the reference time interval and, if the median acceleration value of at least one frequency band of the time interval has been increased by more than 10% compared to the median acceleration at the same frequency band of the reference time interval, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period.

[0023] It is preferred in the aforementioned alternatives ii) to v) that the different frequency bands do not overlap with each other and wherein the gap between adjacent frequency bands, at which the acceleration of the vessel is measured in step b), is less than 100 Hz, preferably less than 10 Hz, more preferably at most 1 Hz and most preferably 0 Hz.

[0024] Also in the above mentioned embodiments it is preferred that it is determined in step b), whether the acceleration of the vessel has been increased within the first predetermined time period - instead of by more than 10% - by at least 20%, more preferably by at least 50%, still more preferably by at least 100% and most preferably by at least 200% of the acceleration of the vessel before the start of the first predetermined time period. In accordance with a further particular preferred embodiment of the present invention, the acceleration values obtained during the time-resolved determination of the acceleration of the vessel in step b) are converted from the time domain into the frequency domain, as described above, with the difference that for each frequency or preferably for each frequency band the power spectral densities (PSD) are determined. The PSD describes the distribution of power into frequency components that make up the acceleration signal and in vibration analysis it is given in the unit of acceleration per Hertz (g2 / Hz). More specifically, the PSD may be visualized as diagram showing for a specific frequency or a specific frequency band or for several subsequent frequency bands each the power distribution of the acceleration signal of the vessel as a function of frequency. In accordance with the present invention, the PSD is calculated as follows: wherein X(f) is the Fourier transform of the time signal x(t) and X*(f) is the complex conjugate of the X(f), where “t” means time and “f” means frequency. Herein, x(t) is the acceleration as a function of time. “Lim Af -> 0” means the limit when the frequency difference Af is made smaller and smaller until it approaches zero.

[0025] In particular, the power spectral density can be derived from the recorded acceleration signal x(t). After the (Fast) Fourier Transform said signal is transferred into the frequency domain depicted by a new variable X(f). The formula shown describes the estimation of the power spectral density form X(f) (complex number) with its complex conjugate X*(f) (same complex number but opposite in sign of the imaginary part) divided by the infinitesimal delta of frequency f. The power spectral density (PSD) is a fundamental concept in signal processing and physics that describes how the power of a signal or time series is distributed across different frequencies. In the context of the present invention, it is in particular utilized to first identify the frequency or frequency band which is most sensitive to the phenomenon to be observed (such as layer formation). Thereby, the recorded and evaluated frequencies to the most sensitive ones to limit the amount of data processed can be used (e.g. measurements up to 10 kHz means over 10000 data points per second which is over 40 kbytes per second or 0.144 Gbytes per hour of raw, unprocessed data) which can provide the most robust signal evaluation. The acceleration can be a direct measurement output by the piezo electronical sensor.

[0026] In this embodiment, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10%, if the power distribution of the acceleration signal of the vessel for at least one specific frequency or for at least one specific frequency band increases within the aforementioned first predetermined time period by more than the first predetermined threshold, i.e. by more than 10%, of the power distribution of the acceleration signal of a reference time interval, i.e. of a time interval ending before the start of the first predetermined time period. The preferred values for the first predetermined time period, for the first predetermined threshold and for the frequency bands of the aforementioned embodiments also apply for this embodiment. Preferably, it is determined in this embodiment in step b), whether the power distribution of the acceleration signal of the vessel for at least one specific frequency or for at least one specific frequency band increases within the aforementioned first predetermined time period by at least 20%, preferably by at least 50%, more preferably by at least 100% and most preferably by at least 200% of the power distribution of the acceleration signal of a reference time interval, i.e. of a time interval ending before the start of the first predetermined time period, such as preferably within the time period having the same length as the first predetermined time period and having been ended at the start of the first predetermined time period. Also in this embodiment it is preferred that the different frequency bands do not overlap with each other and wherein the gap between adjacent frequency bands, at which the accel- eration of the vessel is measured in step b), is less than 100 Hz, preferably less than 10 Hz, more preferably at most 1 Hz and most preferably 0 Hz.

[0027] In a further preferred development of the idea of the present invention, it is proposed that it is determined in step b) whether the power distribution of the acceleration signal of the vessel for the frequency or the frequency band, which has been increased within the aforementioned first predetermined time period by more than 10%, preferably at least 20%, more preferably by at least 50%, still more preferably by at least 100% and most preferably by at least 200% of the power distribution of the acceleration signal of a reference time interval has been decreased within the second predetermined time period by at least 20%, preferably at least 50%, more preferably at least 75% and most preferably at least 90% of the power distribution of the acceleration signal measured within the time period having the same length as the first predetermined time period and having been ended at the start of the first predetermined time period.

[0028] In accordance with a further particular preferred embodiment of the present invention, the acceleration of the vessel is time-resolved determined and recorded at a high frequency in step b) with an accelerometer or vibration sensor, respectively, which is fixed at the shell of vessel. The vibration can be measured with piezoelectric and capacitive units or other sensors based on strain gauges, optical or inductive principles.

[0029] Good results are in particular obtained, when the acceleration of the vessel is time-resolved determined in step b) at at least one location on or within the outer vessel wall. Preferably, the acceleration of the vessel is time-resolved determined in step b) at one to ten, more preferably at one to five, still more preferably at one to three and most preferably at one or two different locations on or within the outer vessel wall. If the vessel is fixed by a frame, rack or other fixation means, it is preferred that the each of the at least one location, at which the acceleration of the vessel is time-resolved determined in step b), is located on or within the vessel wall and has a distance to the frame rack or other fixation means, which holds the vessel, of at least 1 cm and preferably of at least 5 cm. If the acceleration measurement is made to close to the fixation means or even at the fixation means, the acceleration is too low to be reasonably measured.

[0030] As set out above, the present invention is not particularly limited concerning the kind of the at least one agitator, which may be for instance one or more blades, one or more paddles, one or more turbine impellers and / or one or more scraper knives. Most preferably, the vessel comprises as agitator at least one scraper knife being mounted in the vessel so that it contacts or at least comes close during its rotation to the one or more cooled surfaces of the vessel so as to be able to scrape off any deposits, such as crystal layers, from the one or more cooled surfaces of the vessel.

[0031] The method in accordance with the present invention is in particular suitable, when the vessel is a scraped heat exchanger, more preferably a scraped crystallizer and most preferably a scraped suspension crystallizer. The scraped suspension crystallizer is preferably connected with a solid / liquid separator for separating the crystal particles from the suspension. For example, the solid / liquid separator may be a wash column apparatus, which comprise a cylindrical vessel, wherein the cylindrical vessel comprises: i) a piston with a piston head and a piston rod, wherein the piston is arranged reciprocatingly movable in the cylindrical vessel, wherein the piston bounds below the piston head a wash chamber inside the cylindrical vessel and wherein the piston head comprises at least one filter means, ii) an inlet for supplying the crystal suspension into the cylindrical vessel, iii) an outlet for discharging liquid from the cylindrical vessel, iv) an outlet for discharging crystal and / or crystal melt from the cylindrical vessel, v) a circulation conduit for circulat- ing melt arranged outside the cylindrical vessel, which is in communication with the wash chamber, and vi) a means arranged in the wash chamber for restricting the movement of the crystal bed that has been compacted in the wash chamber by the piston and for directing the wash liquid entering into the cylindrical vessel from the circulation conduit so as to homogeneously distribute it over the entire crosssection of the cylindrical vessel. In addition, the scraped suspension crystallizer may be connected with a growth vessel, which comprises at least one agitator, by two lines, wherein one line allows to led suspension from the scraped suspension crystallizer into the interior of the growth vessel and the other line allows to led suspension from the interior of the growth vessel into the interior of the scraped suspension crystallizer, wherein the function of the growth vessel is to allow the crystals being suspended in the liquid to grow.

[0032] The present invention is not particularly restricted concerning the kind of the medium, with which the vessel is filled during the operation of step a). Accordingly, the vessel may be filled during the operation of step a) with any liquid, with any suspension or with any emulsion. In particular, the form of the medium, with which the vessel is filled during the operation of step a) may change over the time. If a scraped crystallizer is used as vessel, the vessel may be filled at the beginning of the operation with a liquid, such as melt or solution, which then converts during the operation due to the formation of crystals within the liquid to a suspension of crystals being suspended in the liquid.

[0033] When it has been determined in step b) that the acceleration of the vessel has been increased within the first predetermined time period by more than the predetermined threshold and preferably when it has been determined in step b) that the acceleration of the vessel has not been decreased within the second predetermined time period by at least 5% of the acceleration of the vessel within the first predetermined time period, a suitable measure may be taken, such as stopping the operation of the vessel and removing the deposition layer from the cooled surface.

[0034] In accordance with a further aspect, the present invention relates to a vessel comprising at least one agitator being particularly suitable for performing the aforementioned method, wherein the vessel comprises a controller being embodied so that it time-resolved determines during the operation of the vessel the acceleration of the vessel and determines, whether the acceleration of the vessel has been increased within a first predetermined time period by more than a predetermined threshold, wherein the controller is embodied so that the length of the first predetermined time period is at most 1 hour, and wherein the predetermined threshold is an increase of the acceleration of the vessel of at least 10% of the acceleration of the vessel before the start of the first predetermined time period.

[0035] Good results are in particular achieved, when the vessel comprises on or within the outer vessel wall at at least one, preferably at one to ten, more preferably at one to five, still more preferably at one to three and most preferably at one or two different locations each an accelerometer or vibration sensor, respectively.

[0036] Furthermore, it is preferred that each of the at least one location, at which the acceleration of the vessel is time-resolved determined in step b), is located on or within the vessel wall and has a distance to an optional frame, which holds the vessel, of at least 1 cm and preferably of at least 5 cm.

[0037] The at least one agitator may be for instance one or more blades, one or more paddles, one or more turbine impellers and / or one or more scraper knives. Most preferably, the vessel comprises as agitator at least one scraper knife being mounted in the vessel so that it contacts or at least come close during its rotation to the one or more cooled surfaces of the vessel so as to be able to scrape off any deposits, such as crystal layers, from the one or more cooled surfaces of the vessel.

[0038] In a further development of the idea of the present invention, it is proposed that the vessel is a scraped heat exchanger, preferably a scraped crystallizer and more preferably a scraped suspension crystallizer.

[0039] In accordance with a further aspect, the present invention relates to a suspension crystallization apparatus, which comprises an aforementioned scraped crystallizer, which is connected via a line with a solid / liquid separator and which is preferably further connected via a line with a growth vessel. Preferably, the solid / liquid separator is a wash column apparatus, which comprise a cylindrical vessel, wherein the cylindrical vessel comprises: i) a piston with a piston head and a piston rod, wherein the piston is arranged reciprocatingly movable in the cylindrical vessel, wherein the piston bounds below the piston head a wash chamber inside the cylindrical vessel and wherein the piston head comprises at least one filter means, ii) an inlet for supplying the crystal suspension into the cylindrical vessel, iii) an outlet for discharging liquid from the cylindrical vessel, iv) an outlet for discharging crystal and / or crystal melt from the cylindrical vessel, v) a circulation conduit for circulating melt arranged outside the cylindrical vessel, which is in communication with the wash chamber, and vi) a means arranged in the wash chamber for restricting the movement of the crystal bed that has been compacted in the wash chamber by the piston and for directing the wash liquid entering into the cylindrical vessel from the circulation conduit so as to homogeneously distribute it over the entire crosssection of the cylindrical vessel.

[0040] The growth vessel preferably comprises at least one agitator and is connected with the scraped crystallizer by two lines, wherein one line allows to led suspension from the scraped suspension crystallizer into the interior of the growth vessel and the other line allows to led suspension from the interior of the growth vessel into the interior of the scraped suspension crystallizer, wherein the function of the growth vessel is to allow the crystals being suspended in the liquid to grow.

[0041] Subsequently, the present invention is described by means of illustrative, but not limiting figures, in which:

[0042] Fig. 1 schematically shows a suspension crystallization apparatus comprising a scraped crystallizer in accordance with one embodiment of the present invention.

[0043] Fig.2 shows a side view of the scraped crystallizer shown in figure 1 .

[0044] Fig. 3 shows the measurement results of example 1 .

[0045] Figure 1 schematically shows a suspension crystallization apparatus 10, which comprises a scraped crystallizer 12, a growth vessel 14 as well as a wash column apparatus 16 as solid / liquid separator. More specifically, the growth vessel 14 is connected with a feed line 18, an inlet line 20 being connected with the scraped crystallizer 12 and an outlet line 22, which leads via a pump 24 into the scraped crystallizer 12. Furthermore, a connection line 26 leads from the scraped crystallizer 12 to the wash column apparatus 16, which comprises a wash column 28, a piston 30, a scraper 32, a filter 34, a piston drive 36, an outlet line 38 for product and an outlet line 40 for residue. In addition, the wash column 28 is connected via a recirculation line 42 via a pump 44 with a heat exchanger 46. In turn, the scraped crystallizer 12 comprises a cooled surface 48, which is cooled by a cooling agent, which is led via the cooling agent inlet line 50 and the cooling agent outlet line 52 across the side of the cooled surface being opposite to the side of the cooled surface being in contact with the liquid contained in the cooled surface of the scraped crystallizer 12. Moreover, the scraped crystallizer 12 comprises a scraper motor 54 as well as several scraper knives 56 being provided within the scraped crystallizer 12. As shown in figure 2, an accelerometer 58 is arranged in the outer wall of the scraped crystallizer 12.

[0046] During the operation of the suspension crystallization apparatus 10, melt to be crystallized is fed via the feed line 18 into the growth vessel 18, in which it is mixed with suspension of crystals being suspended in liquid, which has been formed in the scraped crystallizer 12 and has been transferred into the interior of the growth vessel 18 via line 20. A portion of the mixture is continuously withdrawn and led via line 22 into the scraped crystallizer 12, in which the suspension is cooled via the cooled surface 48 so that crystals of the target compound to be purified are formed in the suspension. The scraper knives 56 are continuously rotated around the cooled surface 48 driven by the scraper motor 54 so as to avoid the formation of crystal layers on the cooled surface 48, which is formed by the inner wall of the vessel of the scraped crystallizer 12. A portion of the suspension is led via line 26 into the wash column 28, in which the crystals are separated from the liquid. During the operation, the acceleration of the vessel of the scraped crystallizer 12 is continuously measured by the accelerometer 58 in short time intervals of for instance 1 millisecond so as to time-resolved determining the acceleration of the vessel during the operation of the vessel and to determine, whether the acceleration of the vessel has been increased within a first predetermined time period by more than a predetermined threshold, wherein the length of the first predetermined time period is at most 1 hour and the predetermined threshold is an increase of the acceleration of the vessel of at least 10% of the acceleration of the vessel before the start of the first predetermined time period. If such an increment of the acceleration is measured, it is concluded that a crystal layer covering about 10% of the cooled surface 48 has been formed on the cooled surface 48 and appropriate measures, such as stopping the process and removing the crystal layer from the cooled surface 48, are taken. Subsequently, the present invention is described by means of an illustrative, but not limiting example.

[0047] Example

[0048] The formation of a deposition layer on a cooled surface being provided inside a vessel was performed using as vessel a scraped suspension crystallizer being embodied as shown in figure 2. As shown in figure 2, a vibration sensor measuring based on a piezoelectric unit was welded onto the center of the wall of the crystallizer and was connected to a personal computer for signal recording. The vibrations signals were collected with a commercial software MatLab, version 2021a) at a rate of 12 kHz and the signals were transformed to the frequency domain using a pre-programmed Fast Fourier Transform (FFT). The generated frequency data were converted into a power spectral density by squaring the fourier-transformed vibration signal and calculating the average value for a certain frequency range.

[0049] A solution of water and 1 % of sodium chloride was prepared for the crystallization process and filled into the scraped suspension crystallizer. Subsequently, all pumps and the scraper were started and the cooling of the crystallizer was initiated. By a slow cooling ramp, crystal formation inside the liquid-phase could be initiated at around -0.8°C liquid-phase temperature without any layer formation, as shown in the section “signal during normal operation” of the diagram of figure 3. After 35 minutes of stable operation, pure water was introduced into the scraped suspension crystallizer via a feed line 18 as shown in figure 1 close to the heat exchanger walls. The local decrease in sodium chloride concentration at constant, low temperature led immediately to a formation of a thin ice layer and was detected by a drastic increase of the power spectral densities of the vessel scraped suspension crystallizer in all observed frequency ranges between 1 and 3000Hz (range intervals 200Hz) as shown at the border between the section “signal during normal operation” and the section “signal after layer formation” in the diagram of figure 3. Subsequently, the ice layer increased slowly in thickness, which can be seen by the positive slope of the power spectral densities.

[0050] Reference Numerals

[0051] Suspension crystallization apparatus

[0052] Scraped crystallizer

[0053] Growth vessel

[0054] Wash column apparatus

[0055] Feed line

[0056] Inlet line

[0057] Outlet line

[0058] Pump

[0059] Connection line

[0060] Wash column

[0061] Piston

[0062] Scraper

[0063] Filter

[0064] Piston drive

[0065] Outlet line for product

[0066] Outlet line for residue

[0067] Recirculation line

[0068] Pump

[0069] Heat exchanger

[0070] Cooled surface

[0071] Cooling agent inlet line

[0072] Cooling agent outlet line

[0073] Scraper motor

[0074] Scraper knife

[0075] Accelerometer

Claims

Sulzer Management AG S13789PEP - Pl / FaClaims:1 . A method for detecting the formation of a deposition layer on a cooled surface being provided inside a vessel comprising at least one agitator, in particular for detecting the formation of a crystal layer on a cooled surface being provided inside a scraped heat exchanger and preferably in a scraped crystallizer, wherein the method comprises the following steps: a) operating the vessel being filled with a liquid, a suspension or an emulsion by driving at least one agitator and by cooling at least one surface and b) time-resolved determining during the operation of the vessel in step a) the acceleration of the vessel and determining, whether the acceleration of the vessel has been increased within a first predetermined time period by more than a predetermined threshold, wherein the length of the first predetermined time period is at most 1 hour and the predetermined threshold is an increase of the acceleration of the vessel of at least 10% of the acceleration of the vessel before the start of the first predetermined time period.

2. The method in accordance with claim 1 , wherein the length of the first predetermined time period in step b) is at most 10 minutes, preferably at most 1 minute, more preferably at most 30 seconds, still more preferably at most 10 seconds, yet more preferably at most 1 second, yet more preferably at most 500 milliseconds and most preferably at most 100 milliseconds of the acceleration of the vessel before the start of the first predetermined time period, wherein the lower boundary for the length of the first predeterminedtime period in step b) is preferably more at least 0.1 milliseconds, more preferably at least 1 millisecond, still more preferably at least 10 milliseconds and most preferably at least 100 milliseconds, wherein preferably the predetermined threshold for the increase of the acceleration of the vessel in step b) is at least 20%, preferably at least 50%, more preferably at least 100% and most preferably at least 200%.

3. The method in accordance with claim 1 or 2, wherein it is further determined in step b), whether the acceleration of the vessel has not been decreased within a second predetermined time period by at least 5% of the acceleration of the vessel within the first predetermined time period, wherein the second predetermined time period starts at the end of the first predetermined time period and the length of the second predetermined time period is at least 100 milliseconds, wherein preferably the length of the second predetermined time period in step b) is at least 250 milliseconds, preferably at least 500 milliseconds, more preferably at least 1 second, yet more preferably at least 2.5 seconds, yet more preferably at least 5 seconds, yet more preferably at least 10 seconds and most preferably at least 20 seconds, wherein the upper boundary for the length of the second predetermined time period in step b) is preferably at most 1 hour, more preferably at most 30 minutes, still more preferably at most 10 minutes, yet more preferably at most 5 minutes and most preferably at most 1 minute, and wherein it is preferably determined in step b), whether the acceleration of the vessel has been decreased within the second predetermined time period by at least 20%, preferably at least 50% and most preferably at least 90% of the acceleration of the vessel within the first predetermined time period.

4. The method in accordance with any of the preceding claims, wherein the acceleration of the vessel is determined in step b) in subsequent time intervals, wherein the numeric values for the acceleration of the vessel measured during the time intervals is at least temporarily stored, wherein each of the time intervals has a length of at most the first predetermined time period, preferably of at most 50% of the first predetermined time period, more preferably of at most 25% of the first predetermined time period and most preferably of at most 10% of the first predetermined time period.

5. The method in accordance with any of the preceding claims, wherein the acceleration values obtained during the time-resolved determination of the acceleration of the vessel in step b) are converted with fast Fourier transformation from the time domain into the frequency domain, before it is determined whether the acceleration of the vessel at a specific frequency or in a specific frequency range has been increased within the first predetermined time period by more than the predetermined threshold.

6. The method in accordance with claim 5, wherein the time-resolved determination of the acceleration of the vessel is measured in step b) at measurement time intervals having a length between 10 milliseconds and 1 hour, preferably between 100 milliseconds and 10 minutes and most preferably between 1 and 10 seconds, wherein subsequent measurement time intervals summing up to time intervals of 10 milliseconds to 1 hour and preferably of 1 second to 1 minute are converted with fast Fourier transformation from the time domain into the frequency domain so as to obtain for subsequent time intervals the frequency dependent acceleration values of the vessel.

7. The method in accordance with claim 5 or 6, wherein the frequency domain comprises frequencies in a range of 1 Hz to 30 kHz.

8. The method in accordance with any of claims 5 to 7, wherein:i) the obtained frequency dependent acceleration values of the vessel for each time interval are compared with the respective frequency dependent acceleration values of the vessel obtained in a reference time interval and it is determined, whether the acceleration at at least one frequency value of the time interval has been increased by more than 10% compared to the acceleration at the same frequency of the reference time interval and, if so, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period, or ii) frequency bands of the subsequent time intervals are compared with each other and it is thereby determined, whether the acceleration at at least one frequency band of a time interval has been increased by more than 10% compared to the acceleration at the same frequency band of a reference time interval and, if so, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period, or iii) average acceleration values of each frequency band are calculated by summing up all acceleration values of the frequency band and by then dividing the sum by the number of frequencies of the frequency band, wherein the average acceleration values of each frequency band of a time interval are compared with the respective average acceleration values of the same frequency band of a reference time interval and, if the average acceleration value of at least one frequency band of a time interval has been increased by more than 10% compared to the average acceleration at the same frequency band of the reference time interval, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than10% of the acceleration of the vessel before the start of the first predetermined time period, or iv) maximal acceleration values of each frequency band are calculated by determining the maximal acceleration values within a frequency band, wherein the highest acceleration value of a frequency band of a time interval is compared with the highest acceleration value of the same frequency band of a reference time interval and, if the highest acceleration value of at least one frequency band of the time interval has been increased by more than 10% compared to the highest acceleration at the same frequency band of the reference time interval, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period, or v) median acceleration values of each frequency band are calculated by determining the median acceleration values within a frequency band, wherein the median acceleration value of a frequency band of a time interval is compared with the median acceleration value of the same frequency band of the reference time interval and, if the median acceleration value of at least one frequency band of the time interval has been increased by more than 10% compared to the median acceleration at the same frequency band of the reference time interval, it is determined that the acceleration of the vessel has been increased within the first predetermined time period by more than 10% of the acceleration of the vessel before the start of the first predetermined time period, wherein preferably in alternatives ii) to iv) the different frequency bands do not overlap with each other and wherein the gap between adjacent frequency bands, at which the acceleration of the vessel is measured in step b), isless than 100 Hz, preferably less than 10 Hz, more preferably at most 1 Hz and most preferably 0 Hz.

9. The method in accordance with any of any of claims 5 to 8, wherein from the frequency domain for each frequency or preferably for each frequency band the power spectral densities are determined, wherein it is determined in step b) that the acceleration of the vessel has been increased within the first predetermined time period by more than 10%, if the power distribution of the acceleration signal of the vessel for at least one specific frequency or for at least one specific frequency band increases within the aforementioned first predetermined time period by more than the first predetermined threshold of the power distribution of the acceleration signal of a reference time interval, wherein the first predetermined threshold is at least 10%, preferably at least 20%, more preferably at least 50%, still more preferably at least 100% and most preferably at least 200%.

10. The method in accordance with claim 9, wherein it is determined in step b) whether the power distribution of the acceleration signal of the vessel for the frequency or the frequency band, which has been increased within the aforementioned first predetermined time period by more than the first predetermined threshold of the power distribution of the acceleration signal before the start of the first predetermined time period, has been decreased within the second predetermined time period by at least 20%, preferably at least 50%, more preferably at least 75% and most preferably at least 90% of the power distribution of the acceleration signal measured within the time period having the same length as the first predetermined time period and having been ended at the start of the first predetermined time period.11 . The method in accordance with any of the preceding claims, wherein the acceleration of the vessel is time-resolved determined and recorded at ahigh frequency in step b) with an accelerometer or vibration sensor, respectively, which is fixed at the shell of vessel, wherein the vibration is preferably measured with piezoelectric, capacitive strain gauges or inductive sensors, wherein preferably each of the at least one location, at which the acceleration of the vessel is time-resolved determined in step b), is located on or within the vessel wall and has a distance to an optional frame, which holds the vessel, of at least 1 cm and preferably of at least 5 cm.

12. The method in accordance with any of the preceding claims, wherein the vessel is a scraped heat exchanger, preferably a scraped crystallizer and more preferably a scraped suspension crystallizer, wherein preferably the vessel is filled in step a) with a liquid, which converts during the operation to a suspension of crystals being suspended in a liquid phase.

13. The method in accordance with any of the preceding claims, wherein the operation of the vessel is stopped and the deposition layer is removed from the cooled surface, when it has been determined in step b) that the acceleration of the vessel has been increased within the first predetermined time period by more than the predetermined threshold and preferably when it has been determined in step b) that the acceleration of the vessel has not been decreased within the second predetermined time period by at least 5% of the acceleration of the vessel within the first predetermined time period.

14. A vessel comprising at least one agitator being particularly suitable for performing a method in accordance with any of the preceding claims, wherein the vessel comprises a controller being embodied so that it time-resolved determines during the operation of the vessel the acceleration of the vessel and determines, whether the acceleration of the vessel has been increased within a first predetermined time period by more than a predeterminedthreshold, wherein the controller is embodied so that the length of the first predetermined time period is at most 1 hour, and wherein the predetermined threshold is an increase of the acceleration of the vessel of at least 10% of the acceleration of the vessel measured within the time period having the same length as the first predetermined time period and having been ended at the start of the first predetermined time period.

15. The vessel in accordance with claim 14, wherein the vessel comprises on or within the outer vessel wall at at least one, preferably at one to ten, more preferably at one to five, still more preferably at one to three and most preferably at one or two different locations each an accelerometer or vibration sensor, respectively, wherein preferably each of the at least one location, at which the acceleration of the vessel is time-resolved determined in step b), is located on or within the vessel wall and has a distance to an optional frame, which holds the vessel, of at least 1 cm and preferably of at least 5 cm.

Images