Bioprocesses process variables

Due to the complexity of bioprocesses, and the lack of direct in-process measurements of critical process variables, much work is being done on development of soft sensors and model predictive control of such systems. Soft sensors have long been used to estimate biomass concentration in fed-batch cultivations. The soft sensors can be integrated into automated control structures to control the biomass growth in the fermentation. 

There are several barriers to the successful control of bioprocesses due to particular circumstances that are related to their characteristics the complexities of microbial metabolisms, the nonlinearity of microbial reactions, the frequent use of batch and fed-batch operations, and the limited availability of sterihzable online sensors for important process variables such as cell and product concentrations. Furthermore, it is difficult to construct mathematical models that can predict the entire range of batch or fed-batch operations that many fermentation processes require. 

These data are utilized for process control, improvement of product quality, the saving of raw material and energy, and an assurance of safety. In this section, we focus on the measurement of process variables as the basis of bioprocess control. 

Process variables measured in bioprocess instrumentation can be categorized into one ofthree groups physical, chemical, and biochemical variables. Table 13.1 summarizes these three types of variables. 

Table 13.1 Process variables measured and estimated in bioprocess instrumentation.

Furthermore, using the primary measurements to obtain the secondary process variables (so-called gateway sensor ) is a form of bioprocess control. For example, the measurement of optical density (primary measurement) can be used for the estimation of cell concentration, and, subsequently, the time course of cell concentration can be employed for the estimation of specific growth rate (secondary... 

The goal of bioprocess control is to maintain important process variables in a bioreactor at a desired level regardless of time-dependent environmental changes. Process control will be performed by the following two steps based on the information obtained through the instrumentation. 

A closed-loop system with feedback, which is illustrated in Figure 13.2, is the central feature of a control system in bioprocess control, as well as in other processing industries. First, a set-point is established for a process variable. Then, the process variable measured in a bioreactor is compared with the set-point value to determine a deviation e. Based on the deviation, a controller uses an algorithm to calculate an output signal O that determines a control action to manipulate a control variable. By repeating this cycle during operation, successful process control is performed. The controller can be the operator when manual control is being employed. 

Figure 13.3 illustrates a typical control action (below) and the response of a process variable (above) in an on - off control system. As a result of on - off control, the response of a process variable inevitably becomes oscillatory around a set-point (overshoot and hunting). Thus, the precise control of a process variable is difficult to achieve using an on-off control system, and, therefore, the on-off control can be applicable only when the minimum and maximum values of the response are acceptable for successful operation of the bioprocess. To decrease the oscillation around a set-point, a differential gap (or a dead zone) is normally used to determine a threshold value. 

Bioprocess Strategy Process variable Control variable Method... 

Figure 2.3. The macroscopic principle applied to bioprocessing expressed with pseudohomogeneous observable process variables in the liquid phase (L) biomass X, substrates S-, oxygen 0, products Pj, carbon dioxide C, and volumetric heat Hy. Pseudohomogeneity is checked by considering a series of mass transfer steps L film at the gas phase-L interface (1), L bulk (2), L film at the L-Solid phase (S) interface (3), cell wall and membranes (4), resp. S-phase cell mass with cytoplasm.

Basics of Quantification Methods for Bioprocesses 19 Table 2.1. Conventional process variables and their measurements. 

Figure 6.51. Diagrammatic representation of a steady-state bioprocess in balance area (reactor) following the macroscopic principle by analyzing elemental composition of significant process variables (substrate, nitrogen source, biomass, product, O2, CO2, H2O). (Adapted from Roels, 1980a.)...

As seen previously for some specific applications such as wastewater treatment plants, software sensors can be envisaged to provide on-line estimation of non-measurable variables, model parameters or to overcome measurement delays [81-83]. Software sensors have been developed mainly for monitoring bioprocesses because the control system design of bioreactors is not straightforward due to [84] significant model uncertainty, lack of reliable on-line sensors, the non-linear and time-varying nature of the system or slow response of the process. 

Some of the earliest work on NIR of bioprocesses was performed on the nutrients and metabolites in a fermentation broth. A classic paper (if 1996 is antiquity) was written by Hall et al.30 on the determination of acetate, ammonia, biomass, and glycerol in E. coli fermentations. This early paper used NIR to simultaneously monitor all the above-mentioned parameters. The correlation coefficients were all better than 0.985 with variable SEPs acetate, 0.7 g/1 ammonia, 7 mM glycerol, 0.7 g/1 and biomass, 1.4 g/1. While later work with more modem equipment has attained better results, this remains as one of the first. The work was performed at line in a cuvette, but rapidly enough to be considered a process measurement. 

The BioView sensor includes a software package (CAMO ASA, Norway) for data analysis and on-Une estimation of different bioprocess variables simultaneously. Thus, the instrument is able to predict the trends of the concentration courses of different variables during a cultivation and is used to give information about important process steps (e.g., feeding time, harvesting time, etc.). The instrument is able to monitor on-line several fluorophores in situ and non-invasively during cultivation processes and permits an estimation of different bioprocess variables simultaneously. The increasing of cell mass concentration and the product formation as well as the actual metabolic state of the cells is simultaneously detectable by this fluorescence technique. 

Obvious goals in bioprocessing are to minimise raw material costs and reaction times, and to maximise product concentrations, yields and purities. However the effect of varying any one variable on the overall process must be accurately assessed. This can be difficult because of the interdependence of many of these variables, making computer models a very useful tool. Thus the use of a cheaper raw material may appear to be attractive, but if this material is of significantly lower purity and so requires costly purification prior to use, or if the product derived from it needs more extensive... 

As with refining and petrochemical processes, bioprocesses must be operated automatically so as to achieve a consistent production of various bioproducts in a cost-effective way. In particular, there is a strong demand to optimize bioprocesses by controlling them automatically to promote labor-saving operations. To achieve this, it is necessary to understand what is happening in a bioreactor (instrumentation) and to properly manipulate the control variables that affect the performance of a bioreactor operation (control). 

The process model can be obtained by different forms, and in bioprocesses mass balance equations canprovide much information. However, in order to have efficient process models and software sensors, a previous adjustment of the model is necessary using on-line data collected from a plant under different operational conditions. This databank is important to guarantee that the model remains calibrated and represents the plant adequately. Some requisites are indispensable for the experimental implementation of models in software sensors response speed to disturbances in the system and appropriate inference of primary variables of interest during key points of the process. 

---