Method for Monitoring Bioprocesses

Abstract

A method of monitoring bioprocesses, wherein a course of a bioprocess is predicted using a process model and values for process parameters are estimated during the bioprocess, where at least one process parameter is selected (step a), for which current measured values are determined during the bioprocess, the respective current measured value of the selected process parameter is compared with the corresponding estimated value for this process parameter estimated by the process model (step b), a variance is then compared with a predetermined threshold value (step c), in a step d), at least one model parameter is then changed when the threshold value is exceeded by the variance, where steps b) to d) are executed until the variance fails to meet the threshold value, and in the case that the threshold value is not met after a predetermined number of repetitions, the method is discontinued and a warning is output.

Claims

1. A method for monitoring bioprocesses, in particular fermentation processes, a course of a bioprocess being predicted aided by a process model and estimated values for biological process parameters being estimated during the course of the bioprocess with the process model, the method comprising: a) selecting at least one process parameter of the bioprocess, for which current measured values are determined during the course of the bioprocess; b) comparing a respective current measured value of the at least one selected process parameter with a corresponding estimated value estimated by the process model for this at least one process parameter; c) comparing a variance between the respective current measured value and the corresponding estimated value for the at least one selected process parameter with a predetermined threshold value; and d) changing at least one defined model parameter employed in the process model when the predetermined threshold value is exceeded by the variance; wherein the steps b) to d) with a respective changed model parameter are executed until the variance is placed below a predetermined threshold value; and wherein in cases that, after a predetermined number of repetitions of steps b) to d) the threshold value is met by the variance, the method is discontinued and a warning is output.

2. The method as claimed in claim 1, wherein a lag term, which is employed as model parameters in the process model, is changed when the threshold value is exceeded within predetermined limit values by the variance during said step b.

3. The method as claimed in claim 2, wherein limit values for changing the lag term are predetermined by a biomass employed in the respective bioprocess, in particular bioactive cells or organisms.

4. The method as claimed in claim 3, wherein the biomass comprises one of bioactive cells and organisms.

5. The method as claimed in claim 1, wherein a process parameter of the bioprocess is selected as a process parameter to be measured, for which during the course of the bioprocess measured values can be determined approximately in real time.

6. The method as claimed in claim 1, wherein the respective repetitions of steps b) to d) are performed with the respective changed model parameter at short intervals in time in relation to one another.

7. The method as claimed in claim 1, wherein at least one of (i) a respiratory quotient, (ii) a concentration of the biomass and (iii) a substrate is selected as a process parameter which is measured during the course of the bioprocess.

8. The method as claimed in one of the preceding claim 1, wherein a deterministic process model is employed to estimate the biological process parameters of the bioprocess.

9. The method as claimed in claim 1, wherein the bioprocesses comprise fermentation processes.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The invention will now be described by way of example, making reference to the accompanying drawing, in which:

[0028] The FIGURE shows a schematic representation of an exemplary course of the inventive method for monitoring bioprocesses, the course of which is predicted with the aid of a process model.

DETAILED DESCRIPTION OF THE EXEMPLARY EMBODIMENTS

[0029] The FIGURE shows a schematic representation of an exemplary course of the inventive method for monitoring bioprocesses based on an exemplary bio or fermentation process with baker's yeast or S. cerevisiae as biomass X. Other bioactive organisms (fungi, bacteria), such as E. coli, or mold, can however also be employed as biomass X in bioprocesses. The bio or fermentation process typically occurs in a bioreactor, in which the environmental and reaction conditions can be optimized and controlled for a bioprocess, such as a fermentation with baker's yeast. To this end, during the bioprocess various environmental and/or process parameters X, S, E, pO2, RQ are regulated and/or measured in the bioreactor such as pH value, temperature, oxygen supply, nitrogen supply, mixer settings, glucose concentration S, concentration of the biomass X or of the substance E to be produced (e.g. ethanol), concentration of dissolved oxygen pO2 or what is known as the respiratory quotient RQ.

[0030] Monitoring of the bioprocess is necessary to maintain and regulate the environmental and reaction conditions. To this end, a process model PM is produced for the course of the bioprocess, with which process model estimated values X.sub.c, S.sub.c, E.sub.c, pO2.sub.c, RQ.sub.c for the respective biological process parameters X, S, E, pO2, RQ, which are relevant to a monitoring and regulation, for instance, can be estimated or calculated for each time segment during the course of the bioprocess. With a fermentation process with baker's yeast, a concentration of biomass C, glucose S, ethanol E, dissolved oxygen pO2 and what are known as respiratory quotients RQ can be determined for the monitoring with the process model PM on the basis of initial values X.sub.0, S.sub.0, E.sub.0, pO2.sub.0. Depending on the respective bioprocess, further and/or other process parameters can also be estimated with the aid of the respective process model PM. Corresponding initial values of the process parameters X, S, E, pO2 of the real bioprocess are typically used as initial values X.sub.0, S.sub.0, E.sub.0, pO2.sub.0 of the process parameters X, S, E, pO2 for the process model PM.

[0031] A deterministic process model PM can be used to estimate the process parameters X, S, E, pO2, RQ for the respective time segment of the bioprocess, for instance, of which deterministic process model PM the knowledge of the respective process-specific, biochemical processes inside and outside of the employed biomass or cells is translated into mathematical equations during the course of the bioprocess. Furthermore, the process model PM can contain at least one defined model parameter with an initial value .sub.0. This model parameter may be what is known as a lag term , for instance, which is used to take into account that the bioprocess or the fermentation can occur more quickly or slowly using baker's yeast depending on the operator, cell strain of the biomass C, preparatory culture of the biomass X, on account of fluctuations in the substrate S, etc., in particular in spite of otherwise identical conditions.

[0032] The method in accordance with the invention begins with a start step, with which the bioprocess or the fermentation with baker's yeast is started in the bioreactor by adding the biomass X to the substrate or glucose S, for instance. The calculation or estimation of the process parameters X, S, E, pO2, RQ for each time segment of the bioprocess is started in parallel with the starting step based on the initial values X.sub.0, S.sub.0, E.sub.0, pO2.sub.0 of the process parameters X, S, E, pO2 by the process model PM. Bioprocess or fermentation and process model are largely to proceed in parallel. At the starting step, the initial value .sub.0 of the model parameter or lag term is employed in the process model PM and the process parameters X, S, E, pO2, RQ are estimated accordingly.

[0033] In a first method step a), at least one process parameter X, S, E, pO2, RQ of the bioprocess is selected. During the course of the bioprocess current measured values RQ.sub.m are determined for this selected process parameter RQ. Ideally the measured values RQ.sub.m of the selected process parameter RQ can be determined approximately in real time during the course of the bioprocess. In the present exemplary course of the method for a fermentation with baker's yeast, what is known as the respiratory quotient RQ was selected as the process parameter RQ.sub.m to be measured, for instance. The respiratory quotient RQ describes a ratio of the carbon dioxide quantity (CO.sub.2) produced at a given time to the simultaneously consumed oxygen (O.sub.2) and represents an indicator of processes within a bioactive cell. During the bioprocess or during the fermentation the process parameter RQ can be measured relatively easily in real time, such as using what is known as off-gas analysis.

[0034] Alternatively or in addition, in the first method step, a concentration of the biomass X and/or a concentration of the glucose S can also be selected with the exemplary bioprocess (fermentation with baker's yeast), as the process parameter X.sub.m, S.sub.m to be measured. The respective concentration of the biomass X can be measured during the course of the fermentation on the basis of the electrical properties of the cells or the biomass X, for instance. The respective concentration of the glucose S can be currently determined, e.g., with the aid of spectroscopy, in particular reflection spectroscopy, based on spectroscopic properties of the substrate during the course of the fermentation.

[0035] In a second method step b), a current measured value RQ.sub.m of the selected process parameter RQ is then compared with the corresponding estimated value RQ.sub.c of the selected process parameter RQ estimated by the process model PM. This can be performed approximately in real time when the bioprocess and process model are run in parallel. That is, a current measured value RQ.sub.m for the respiratory quotient RQ and an estimated value RQ.sub.c determined for the measuring time instant by the process model PM are determined as synchronously as possible for the respiratory quotient RQ during the course of the fermentation and then compared with one another. The comparison can be realized as subtracting the estimated value RQ.sub.c from the measured value RQ.sub.m or as subtracting the measured value RQ.sub.m from the estimated value RQ.sub.c. If the measured value RQ.sub.m and the estimated value RQ.sub.c for the respective time instant are not identical, then a variance RQ is provided by the subtraction.

[0036] In a third method step c), the variance RQ between the respective current measured value RQ.sub.m and the corresponding estimated value RQ.sub.c is compared with a predetermined threshold value. On account of the comparison via subtraction the variance RQ between measured value RQ.sub.m and estimated value RQ.sub.c can have a positive or negative sign. As a result, a sum of the variance RQ determined in the second method step b) is used for the comparison with the predetermined threshold value.

[0037] If it is established in the third method step c) that the variance RQ or the sum of the variance RQ is less than the predetermined threshold value, then the process model PM is classified as a good prediction of the course of the bioprocess or fermentation. That is, it is assumed therefrom that valid estimated values X.sub.c, S.sub.c, E.sub.c, pO2.sub.c of the further process parameters X, S, E, pO2 are provided by the process model PM and in the course of the bioprocess or fermentation these estimated values X.sub.c, S.sub.c, E.sub.c, pO2.sub.c can be used for a monitoring. The inventive method is ended with the validation of the used process model PM.

[0038] If it is established in the third method step c) that the variance RQ or the sum of the variance RQ has exceeded the predetermined threshold value, then a fourth method step d) is performed. In the fourth method step d), the defined model parameter employed in the process model PM, such as the lag term , changes within predetermined limit values. The limit values for this change in the model parameter are predetermined, for instance, by the biomass X used in the bioprocess. In the present example, a lag term is employed as the model parameter for the inventive method and baker's yeast or the bioactive organism S. cerevisiae is used as biomass X for the bioprocess or fermentation. This thus also specifies the limit values for the variation of the model parameter or lag term . This means that in the fourth method step d), the model parameter or the lag term of the process model PM is changed from an initial value .sub.0 to a new value .sub.1, where the new value .sub.1 for the lag term has to lie within the predetermined limit values.

[0039] Subsequently the second and the third method step b), c) are run through again, where the new estimated value RQ.sub.c of the selected process parameter RQ is calculated for the now current time segment by the bioprocess of the process model PM with the changed value .sub.1 of the model parameter or lag term . This new estimated value RQ.sub.c is then compared with a now current measured value RQ.sub.m of the selected process parameter RQ in the repetition of the second method step b). In the repetition of the third method step c), the variance RQ that is determined based on the now current values for the measured value RQ.sub.m and the estimated value RQ.sub.c or the sum of this variance RQ is compared again with the predetermined threshold value. If the variance RQ or the sum of the variance RQ is now less than the predetermined threshold value, then the original variance RQ that is above the threshold value and, which has been established during a first or preceding run of the second and third method step b), c), is assessed as a biological variability or as a variability in the process control for instance. The bioprocess or the fermentation can thus be continued and the predictions X.sub.c, S.sub.c, E.sub.c, pO2.sub.c of the further process parameters X, S, E, pO2 can thus be classified as validated. The method is then terminated.

[0040] If in the repetition of the third method step c) the threshold value is again exceeded by the newly determined variance RQ, then the fourth method step d) can be performed again. That is, the model parameter or the lag term can be again varied within the predetermined limit values. The second to fourth method steps b), c), d) can then be run through with a respective modified model parameter or lag term for the process model PM until the respective determined variance RQ is placed below the predetermined threshold value. With this variance RQ, the predictions of the further process parameters X, S, E, pO2 can then be considered as validated and the method can be terminated. The established variances RQ can be classified again as biological variability or variability which barely influence the bioprocess.

[0041] Ideally the respective repetitions of the second to fourth method steps b), c), d) are performed with the respective modified model parameter or delay term for the process model PM in short intervals in time (e.g., 10 seconds), in order quickly to arrive at a validated process model PM or at a high-quality classification of the detected variances RQ.

[0042] In the case that after a predetermined number of repetitions of the second to fourth method step b), c), d) with a respective changed model parameter or lag term for the process model PM, the predetermined threshold value of the respective detected variance RQ between the respective current measured value RQ.sub.m and the corresponding estimated value RQ.sub.c of the selected process parameter RQ is still exceeded, the method is discontinued. A warning to the operator can then be output, for instance, in order to effect an intervention in the bioprocess or the fermentation. With this intervention, the bioprocess can be assessed and, e.g., the sensor technology monitored. In the worst case the bioprocess must be terminated.

[0043] In the case that the variance RQ cannot be eliminated by repeatedly varying the model parameter , this variance RQ is considered to be a process variance or it is assumed herefrom that there is an error in the process, such as a failure of the pH value control, a faulty or incorrect gassing, or contamination in the reactor. The estimated values X.sub.c, S.sub.c, E.sub.c, pO2.sub.c of the further process parameters X, S, E, pO2 which are determined with the process model PM can thus be classified as not validated. By repeating the second to fourth method step b), c), d), with the respective modified model parameter or lag term for the process model PM at short intervals in time (e.g. 10 seconds), in this case a near-instant and quick reaction can be made. Possible damage to the bioprocess and/or equipment can thus be avoided.

[0044] On the whole the inventive method is advantageous in that a number of bioprocesses, which in case of doubt are otherwise discontinued, is reduced, so that a developing variance RQ between the respective current measured value RQ.sub.m and the corresponding estimated value RQ.sub.c of the selected process parameter RQ can be evaluated and classified in terms of quality. As a result, the efficiency in the use of resources and raw materials can be increased and the operating costs can be minimized.

[0045] Thus, while there have been shown, described and pointed out fundamental novel features of the invention as applied to a preferred embodiment thereof, it will be understood that various omissions and substitutions and changes in the form and details of the devices illustrated, and in their operation, may be made by those skilled in the art without departing from the spirit of the invention. For example, it is expressly intended that all combinations of those elements and/or method steps which perform substantially the same function in substantially the same way to achieve the same results are within the scope of the invention. Moreover, it should be recognized that elements and/or method steps shown and/or described in connection with any disclosed form or embodiment of the invention may be incorporated in any other disclosed or described or suggested form or embodiment as a general matter of design choice. It is the intention, therefore, to be limited only as indicated by the scope of the claims appended hereto.