A Liquid Food Product is Processed in a Continuous Flow Sterilizer

Continuous Sterilisation

Continuous sterilisation then, is the epitome of the HTST method where the nutrient degradation is minimised for any specified degree of sterilisation.

From: Bioprocess Engineering , 2013

Reactor Engineering

Pauline M. Doran , in Bioprocess Engineering Principles (Second Edition), 2013

14.6.2 Continuous Heat Sterilisation of Liquids

Continuous sterilisation, particularly a high-temperature, short-exposure-time process, can reduce thermal damage to the medium significantly compared with batch sterilisation, while achieving high levels of cell destruction. Other advantages include improved steam economy and more reliable scale-up. The amount of steam needed for continuous sterilisation is 20 to 25% of that used in batch processes; the time required is also significantly reduced because heating and cooling are virtually instantaneous.

Typical equipment configurations for continuous sterilisation are shown in Figure 14.38. In Figure 14.38(a), raw medium entering the system is first preheated by hot, sterile medium in a heat exchanger; this reduces the subsequent steam requirements for heating and cools the sterile medium. Steam is then injected directly into the medium as it flows through a pipe; as a result, the temperature of the medium rises almost instantaneously to the desired sterilisation temperature. The time of exposure to this temperature depends on the length of pipe in the holding section of the steriliser. After sterilisation, the medium is cooled instantly by flash cooling, which is achieved by passing the liquid through an expansion valve into a vacuum chamber. Further cooling takes place in the heat exchanger where residual heat is used to preheat incoming medium. The sterile medium is then ready for use in the fermenter.

Figure 14.38. Continuous sterilising equipment: (a) continuous steam injection with flash cooling; (b) heat transfer using heat exchangers.

Figure 14.38(b) shows an alternative sterilisation scheme based on heat exchange between the medium and steam. Raw medium is preheated using hot, sterile medium in a heat exchanger and is brought to the sterilisation temperature by further heat exchange with steam. The sterilisation temperature is maintained in the holding section; the sterile medium is then cooled by heat exchange with incoming medium before being used in the fermenter. Heat exchange systems are more expensive to construct than injection devices; fouling of the internal heat exchange surfaces (Section 9.4.4) also reduces the efficiency of heat transfer between cleanings. On the other hand, a disadvantage associated with steam injection is dilution of the medium by condensate; foaming from direct steam injection can also cause problems with operation of the flash cooler. As indicated in Figure 14.39, the rates of heating and cooling in continuous sterilisers are much more rapid than in batch systems (Figure 14.36). Accordingly, in the design of continuous sterilisers, contributions to cell death outside of the holding period are generally ignored.

Figure 14.39. Variation of temperature with time in the continuous sterilisers of Figure 14.38: (a) continuous steam injection with flash cooling; (b) heat transfer using heat exchangers.

An important variable affecting the performance of continuous sterilisers is the nature of the fluid flow in the system. Ideally, all fluid entering the equipment at a particular instant should spend the same time in the steriliser and exit the system at the same time. Unless this occurs, we cannot fully control the time spent in the steriliser by all fluid elements. No mixing should take place in the tubes: if fluid nearer the entrance of the pipe mixes with fluid ahead of it, there is a risk that contaminants will be transferred to the outlet of the steriliser. The type of flow in pipes where there is neither mixing nor variation in fluid velocity is called plug flow, as already described in Section 14.5.8 for plug flow reactors. Plug flow is an ideal flow pattern: in reality, fluid elements in pipes have a range of different velocities. As illustrated in Figure 14.40, flow tends to be faster through the centre of pipes than near the walls. Plug flow is approached in pipes at high Reynolds numbers (Section 7.2.3) above about 2×10⁴. However, although operation at high Reynolds number is used to minimise fluid mixing and velocity variation in continuous sterilisers, deviations from ideal plug flow are inevitable.

Figure 14.40. Velocity distributions for flow in pipes. (a) In plug flow, the fluid velocity is the same across the diameter of the pipe as indicated by the arrows of equal length. (b) In fully-developed turbulent flow, the velocity distribution approaches that of plug flow; however there is some reduction of flow speed at the walls. (c) In laminar flow, the fluid velocity is lowest at the walls of the pipe and highest along the central axis of the tube.

Deviation from plug flow behaviour is characterised by axial dispersion in the system, that is, the degree to which mixing occurs along the length or axis of the pipe. Axial dispersion is a critical factor affecting the design of continuous sterilisers. If axial dispersion is substantial, the performance of the steriliser is reduced. The relative importance of axial dispersion and bulk flow in the transfer of material through pipes is represented by a dimensionless variable called the Peclet number:

(14.101) $Pe=\frac{uL}{{\mathcal{D}}_z}$

where Pe is the Peclet number, u is the average linear fluid velocity, L is the pipe length, and ${\mathcal{D}}_z$ is the axial dispersion coefficient. For perfect plug flow, ${\mathcal{D}}_z$ is zero and Pe is infinitely large. In practice, Peclet numbers between 3 and 600 are typical. The value of ${\mathcal{D}}_z$ for a particular system depends on the Reynolds number and pipe geometry; a correlation from the engineering literature for evaluating ${\mathcal{D}}_z$ is shown in Figure 14.41.

Figure 14.41. Correlation for determining the axial dispersion coefficient in turbulent pipe flow. Re is the Reynolds number, D is the pipe diameter, u is the average linear fluid velocity, ρ is the fluid density, μ is the fluid viscosity, and ${\mathcal{D}}_z$ is the axial dispersion coefficient. Data were measured using single fluids in: (●) straight pipes; (■) pipes with bends; (□) artificially roughened pipe; and (○) curved pipe.

Reprinted (adapted) with permission from O. Levenspiel, Longitudinal mixing of fluids flowing in circular pipes. Ind. Eng. Chem. 50, 343–346. Copyright 1958, American Chemical Society.

Once the Peclet number has been calculated from Eq. (14.101), the extent of cell destruction in the steriliser can be related to the cell specific death constant k _d using Figure 14.42. In this figure, N ₁ is the number of viable cells entering the steriliser, N ₂ is the number of cells leaving, Pe is the Peclet number as defined by Eq. (14.101), and Da is another dimensionless number called the Damköhler number:

Figure 14.42. Thermal destruction of contaminating organisms as a function of the Peclet number Pe and Damköhler number Da. N ₁ is the number of viable cells entering the holding section of the steriliser; N ₂ is the number of cells leaving.

From S. Aiba, A.E. Humphrey, and N.F. Millis, 1965, Biochemical Engineering, Academic Press, New York.

(14.102) $Da=\frac{k_{\mathrm{d}}L}{u}$

where k _d is the specific death constant, L is the length of the holding pipe, and u is the average linear liquid velocity. The relationship between k _d and temperature is given by Eq. (12.74). The lower the value of N ₂/N ₁, the greater is the level of cell destruction. Figure 14.42 shows that, at any given sterilisation temperature defining the value of k _d and therefore Da, the performance of the steriliser declines significantly as the Peclet number decreases, reflecting the detrimental effect of axial dispersion on steriliser efficiency. Design calculations for a continuous steriliser are illustrated in Example 14.8.

Example 14.8

Holding Temperature in a Continuous Steriliser

Liquid medium at a flow rate of 2   m³  h^–1 is to be sterilised by heat exchange with steam in a continuous steriliser. The medium contains bacterial spores at a concentration of 5×10¹²  m^–3. Values of the activation energy and Arrhenius constant for thermal destruction of these contaminants are 283   kJ   gmol^–1 and 5.7×10³⁹  h^–1, respectively. A contamination risk of one organism surviving every 60 days of operation is considered acceptable. The steriliser pipe has an inner diameter of 0.1   m and the length of the holding section is 24   m. The density of the medium is 1000   kg   m^–3 and the viscosity is 3.6   kg   m^–1  h^–1. What sterilising temperature is required?

Solution

The desired level of cell destruction is evaluated using a basis of 60 days. Ignoring any cell death in the heating and cooling sections, the number of cells entering the holding section over 60 days is:

$N_1=2{\mathrm{m}}^3{\mathrm{h}}^{-1}(5\times {10}^{12}{\mathrm{m}}^{-3})\cdot \left|\frac{24\mathrm{h}}{1\mathrm{day}}\right|\cdot (60\mathrm{days})=1.44\times {10}^{16}$

N ₂, the acceptable number of cells leaving during this period, is 1. Therefore:

$\frac{N_2}{N_1}=\frac{1}{1.44\times {10}^{16}}=6.9\times {10}^{-17}$

The linear velocity u in the steriliser is equal to the volumetric flow rate divided by the cross-sectional area of the pipe:

$u=\frac{2{\mathrm{m}}^3{\mathrm{h}}^{-1}}{\pi {\left(\frac{0.1\mathrm{m}}{2}\right)}^2}=254.6\mathrm{m}{\mathrm{h}}^{-1}$

Calculating the Reynolds number for pipe flow using Eq. (7.1):

$Re=\frac{Du\rho }{\mu }=\frac{(0.1\mathrm{m})(254.6\mathrm{m}{\mathrm{h}}^{-1})(1000\mathrm{kg}{\mathrm{m}}^{-3})}{3.6\mathrm{kg}{\mathrm{m}}^{-1}{\mathrm{h}}^{-1}}=7.07\times {10}^3$

For this value of Re, we can determine ${\mathcal{D}}_z$ from Figure 14.41 using either the experimental or theoretical curve. Let us choose the experimental curve as this gives a larger value of ${\mathcal{D}}_z$ and a smaller value of Pe; the steriliser design will thus be more conservative. Therefore, for ${\mathcal{D}}_z/uD=0.65$ :

${\mathcal{D}}_z=0.65(254.6\mathrm{m}{\mathrm{h}}^{-1})(0.1\mathrm{m})=16.5{\mathrm{m}}^2{\mathrm{h}}^{-1}$

From Eq. (14.101):

$Pe=\frac{uL}{{\mathcal{D}}_z}=\frac{(254.6\mathrm{m}{\mathrm{h}}^{-1})(24\mathrm{m})}{16.5{\mathrm{m}}^2{\mathrm{h}}^{-1}}=370$

Using Figure 14.42, we can determine the value of k _d for the desired level of cell destruction. For N ₂/N ₁=6.9×10^–17 and Pe=370, the corresponding value of Da is about 42. Therefore, from Eq. (14.102):

$k_{\mathrm{d}}=\frac{uDa}{L}=\frac{(254.6\mathrm{m}{\mathrm{h}}^{-1})(42)}{24\mathrm{m}}=445.6{\mathrm{h}}^{-1}$

The sterilisation temperature can be evaluated from the Arrhenius equation after rearranging Eq. (12.74). Dividing both sides by A and taking natural logarithms gives:

$\ln \frac{k_{\mathrm{d}}}{A}=\frac{-E_{\mathrm{d}}}{RT}$

Therefore:

$T=\frac{\left(\frac{-E_{\mathrm{d}}}{R}\right)}{\ln \left(\frac{k_{\mathrm{d}}}{A}\right)}$

E _d=283   kJ   gmol^–1=283×10³  J   gmol^–1; A=5.7×10³⁹  h^–1. From Appendix B, the ideal gas constant R is 8.3144   J   K^–1  gmol^–1. Therefore:

$T=\frac{\left(\frac{-283\times {10}^3\mathrm{J}{\mathrm{gmol}}^{-1}}{8.3144\mathrm{J}{\mathrm{K}}^{-1}{\mathrm{gmol}}^{-1}}\right)}{\ln \left(\frac{445.6{\mathrm{h}}^{-1}}{5.7\times {10}^{39}{\mathrm{h}}^{-1}}\right)}=398.4\mathrm{K}$

Using the conversion between K and °C given in Eq. (2.27), T=125°C. Therefore, the sterilisation temperature required is 125°C.

Heating and cooling in continuous sterilisers are so rapid that in design calculations they are considered instantaneous. While reducing nutrient deterioration, this feature of the process can cause problems if there are solids present in the medium. During heating, the temperature at the core of solid particles remains lower than in the medium. Because of the extremely short contact times in continuous sterilisers compared with batch systems, there is a much greater risk that particles will not be heated thoroughly and will therefore not be properly sterilised. It is important that raw medium be clarified as much as possible before it enters a continuous steriliser.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780122208515000149

Bioprocess asepsis and sterility

Kim Gail Clarke , in Bioprocess Engineering, 2013

10.1.3 Continuous sterilisation

Continuous sterilisers consist of a single pipe or multiple pipes in parallel through which the medium is pumped at the sterilisation temperature. Heating is accomplished either by direct steam injection or by indirect heat exchange.

In the case of continuous sterilisation by direct steam injection, the heating of the medium to the sterilisation temperature is almost instantaneous. Cooling of the hot medium after sterilisation takes place by flash vaporisation, similarly almost instantaneously.

Indirect heat exchange is typically carried out by two heat exchangers in series on each end of the steriliser. The medium at room temperature is heated initially via heat exchange with the hot medium exiting the steriliser, and subsequently via heat exchange with steam to the sterilisation temperature. This second heat exchanger also serves to cool the exiting hot sterile medium which, if necessary, undergoes further heat exchange with cooling water to the operating temperature.

Both the direct and indirect methods provide such fast heating and cooling that the heating and cooling periods can be neglected in terms of their effect on both thermal death and thermal degradation. Under continuous sterilisation conditions then, the time of heating and cooling is insignificant and the time at which the medium is held at the sterilisation temperature constitutes the total sterilisation time.

Continuous sterilisation then, is the epitome of the HTST method where the nutrient degradation is minimised for any specified degree of sterilisation. This is the basis of the process for milk sterilisation which extends the shelf-life considerably further than that of pasteurised milk. ⁹ Typically the milk is held at a temperature in the range of 130°C to 150°C for up to 30 seconds during which time the sterilisation takes place. At this temperature, the thermal degradation rate constant is much lower than the thermal death rate constant, comparatively little degradation takes place in this short time and the nutrient value of the milk is maintained. ¹⁰ A small degree of caramelisation does take place, however, resulting in a slightly altered taste from that of pasteurised milk.

During continuous sterilisation, as with batch sterilisation, the total sterilisation time required to reach the specified degree of sterilisation needs to be quantified. This is readily calculated from the ratio of the total pipe length to the velocity of the medium in the pipes. However, fluid flow in pipes experiences axial dispersion which means that the velocity of the medium in the pipes will differ depending on its radial position in the pipe. Maximal velocity will be experienced at the centre of the pipe but will decrease with position towards the pipe wall. This velocity gradient means that the medium near the centre of the pipe will have a shorter than average residence time in the pipe while the medium closer to the edges will have a longer than average residence time in the pipe. Consequently the design cannot simply be based on an average velocity because the elements near the centre will not be sufficiently sterilised while those at the edges will be over sterilised.

Clearly a design which does not enable the fluid elements at the centre to be subjected to the required sterilisation time is flawed as the specified sterilisation criterion will not be met. However, the option of increasing the sterilisation time to account for the fluid elements moving at the maximum velocity is likewise flawed. Under these circumstances, all fluid elements moving with a velocity lower than average will be held at the sterilisation temperature for a time longer than is necessary to effect the specified degree of sterilisation. Thus all these fluid elements would have undergone a degree of thermal degradation above that which is necessary for efficient sterilisation.

In order to ensure efficient sterilisation of all the fluid elements, while at the same time minimising the effects of thermal degradation, it is imperative that the magnitude of the axial dispersion be quantified and taken into account in the calculations. This is particularly important at Reynolds numbers associated with laminar flow where axial dispersion results in a parabolic velocity profile with a maximum velocity twice that of the average. In turbulent flow, the effect of axial dispersion is less and the velocity profile is flattened. As plug flow is approached, the maximum velocity is only about 1.2-fold of the average velocity.

To account for the variations in sterilisation times, a dispersion model is used. The model is developed by executing a mass balance on a differential fluid volume, of length dx, in continuous steriliser with a uniform cross- sectional area. Microorganisms enter and leave the differential volume by convective flow and by diffusive (axial dispersion) flux. During the time spent in the differential volume, the number of viable microorganisms decreases by dN. The process is assumed to take place at steady state (i.e. no accumulation in the system) and the mass balance solved with the appropriate boundary conditions. The solution to the model is organised in terms of the dimensionless Péclet (Pe) ¹¹ and Damköhler (Da) ¹² numbers where ${\mathrm{Pe}=\overline{\mathrm{u}}\mathrm{L}/\mathrm{D}}_z$ ¹³ and $\mathrm{Da}={\mathrm{k}}_{\mathrm{d}}\mathrm{L}/\overline{\mathrm{u}}$ (Equation 10.14), where δ is given by Equation 10.15. ¹⁴ This equation predicts the fractional degree of sterilisation $\left(\frac{\mathrm{N}}{{\mathrm{N}}_0}\right)$ as a function of the extent of cell destruction (Da) and the extent of axial dispersion (Pe).

[10.14] $\frac{\mathrm{N}}{{\mathrm{N}}_0}=\frac{4\delta {\mathrm{e}}^{\mathrm{Pe}/2}}{{\left(1+\delta \right)}^2{\mathrm{e}}^{\delta \mathrm{Pe}/2}-{\left(1-\delta \right)}^2{\mathrm{e}}^{-\delta \mathrm{Pe}/2}}$

[10.15] $\delta =\sqrt{1+\frac{4\mathrm{Da}}{\mathrm{Pe}}}$

When axial dispersion is negligible, D_z → 0, Pe → ∞ and perfect plug flow is approached. Under these conditions, Equation 10.14 reduces to Equation 10.16 where $\mathrm{L}/\overline{\mathrm{u}}$ is the average residence time in the pipe. Thus this equation predicts the degree of sterilisation based purely on the average residence time and is analogous to that predicting the degree of sterilisation during the holding time in a batch sterilisation process (Equation 10.8). In reality, however, Pe falls mainly between 10 and 1000 and real flow behaviour has to be taken into account.

[10.16] $\frac{\mathrm{N}}{{\mathrm{N}}_0}={\mathrm{e}}^{-\mathrm{Da}}={\mathrm{e}}^{-{\mathrm{k}}_{\mathrm{d}}\mathrm{L}/\overline{\mathrm{u}}}$

The solution to the equation is traditionally depicted graphically on a semi log scale with $\frac{\mathrm{N}}{{\mathrm{N}}_0}$ on the y-axis, Da on the x-axis and Pe as the third parameter (Figure 10.3). This graphical representation shows that as plug flow is approached (Pe → ∞) the length of piping, and hence the residence time, decreases for the same degree of sterility. Further, as the degree of sterility required increases, the effect of the deviation from plug flow on the increase in length of piping, or residence time, is enhanced. Since very high degrees of sterilisation are normally required (say 99%), this fact highlights the importance of the incorporation of the axial dispersion of flow on the design of the steriliser.

Figure 10.3. Thermal destruction of contaminating microorganisms as a function of the extent of destruction and extent of axial dispersion

(Reprinted from Biochemical Engineering, 2nd edition, Alba, Humphrey and Millis, University of Tokyo Press (1973))

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9781782421672500107

Engineering Fundamentals of Biotechnology

M. Berovič , in Comprehensive Biotechnology (Second Edition), 2011

2.12.3.3.5 Main advantages of continuous sterilization

The main advantages of continuous sterilization compared to the batch sterilization are safer sterilization and lower and more uniform demand on services. The design will lead to a safe control of sterilization and elimination of cold pockets where the bioprocess broth could be insufficiently treated. This will result in a safer sterilization and the treated broth can be made much more consistent in quality than that from a batch type.

Continuous sterilization of the broth means a total heat treatment of only about 3–5   min, compared to the batch one where for the same volume a total time of 3–10   h is needed. The more latent heat treatment of the broth can result in a higher yield, especially for heat-sensitive types of broth [3, 6].

Batch sterilization needs large amounts of steam and cooling water for heating and cooling the liquid when the broth is heated at the sterilization and cooled to the bioprocess temperature, respectively.

These peaks in consumption can be cut off by the use of continuous sterilization. This means that for a new installation a considerably smaller steam boiler is needed for a certain bioprocess capacity. Alternatively, increased production capacity can be achieved in the existing plant with the same steam boiler by the introduction of continuous sterilization. The possibility of recovering heat can reduce the consumption of steam and cooling water up to 60–80% [13].

With batch sterilization, the broth is put into the bioreactor mixed and then sterilized and cooled on the same place. The heating and cooling periods are very time consuming, especially for the large units. With continuous sterilization, first, the empty bioreactors are sterilized and, second, the already sterilized and cooled product is fed into it. This means that the preparation time can be shortened and thus the bioreactor can be more efficiently used [18].

Disadvantages of this process are only salt precipitation due to rapid heating and cooling, increase of viscosity by using starch or other polymers and, sometimes, instability due to bigger particles.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780080885049000933

Sterilization

Peter F. Stanbury , ... Stephen J. Hall , in Principles of Fermentation Technology (Third Edition), 2017

Design of continuous sterilization processes

The design of continuous sterilization cycles may be approached in exactly the same way as for batch sterilization systems. The continuous system includes a time period during which the medium is heated to the sterilization temperature, a holding time at the desired temperature, and a cooling period to restore the medium to the fermentation temperature. The temperature of the medium is elevated in a continuous heat exchanger and is then maintained in an insulated serpentine holding coil for the holding period. The length of the holding period is dictated by the length of the coil and the flow rate of the medium. The hot medium is then cooled to the fermentation temperature using two sequential heat exchangers—the first utilizing the incoming medium as the cooling source (thus conserving heat by heating-up the incoming medium) and the second using cooling water. As discussed earlier, the major advantage of the continuous process is that a much higher temperature may be utilized, thus reducing the holding time and reducing the degree of nutrient degradation. The required Del factor may be achieved by the combination of temperature and holding time that gives an acceptably small degree of nutrient decay. Richards (1968) quoted the following example to illustrate the range of temperature-time regimes that may be employed to achieve the same probability of obtaining sterility. The Del factor used by Richards for the example sterilization was 45.7 and the following temperature time regimes were calculated such that they all gave the same Del factor value of 45.7:

Temperature (°C)	Holding time
130	2.44 min
135	51.9 s
140	18.9 s
150	2.7 s

Furthermore, because a continuous process involves treating small increments of medium, the heating-up and cooling-down periods are very small compared with those in a batch system. There are two types of continuous sterilizer that may be used for the treatment of fermentation media: the indirect heat exchanger and the direct heat exchanger (steam injector).

The most suitable indirect heat exchangers are of the double-spiral type, which consists of two sheets of high-grade stainless steel, which have been curved around a central axis to form a double spiral, as shown in Fig. 5.11. The ends of the spiral are sealed by covers. A full-scale example is shown in Fig. 5.12. To achieve sterilization temperatures steam is passed through one spiral from the center of the exchanger and medium through the other spiral from the outer rim of the exchanger, in countercurrent streams. Spiral heat exchangers are also used to cool the medium after passing through the holding coil. Incoming unsterile medium is used as the cooling agent in the first cooler so that the incoming medium is partially heated before it reaches the sterilizer and, thus, heat is conserved.

Figure 5.11. A Schematic Representation of a Spiral Heat Exchanger

Alfa-Laval Engineering Ltd, Camberley, Surrey, UK

Figure 5.12. Industrial Scale Spiral Heat Exchanger

Alfa-Laval Engineering Ltd., Camberley, Surrey, UK

The major advantages of the spiral heat exchanger are:

1.

The two streams of medium and cooling liquid, or medium and steam, are separated by a continuous stainless steel barrier with gasket seals being confined to the joints with the end plates. This makes cross contamination between the two streams unlikely.

2.

The spiral route traversed by the medium allows sufficient clearances to be incorporated for the system to cope with suspended solids. The exchanger tends to be self-cleaning which reduces the risk of sedimentation, fouling, and "burning-on."

Indirect plate heat exchangers consist of alternating plates through which the countercurrent streams are circulated. Gaskets separate the plates and failure of these gaskets can cause cross-contamination between the two streams. Also, the clearances between the plates are such that suspended solids in the medium may block the exchanger and, thus, the system is only useful in sterilizing completely soluble media. However, the plate exchanger is more adaptable than the spiral system in that extra plates may be added to increase its capacity.

The continuous steam injector injects steam directly into the unsterile broth. The advantages and disadvantages of the system have been summarized by Banks (1979):

1.

Very short (almost instantaneous) heating up times.

2.

It may be used for media containing suspended solids.

3.

Low capital cost.

4.

Easy cleaning and maintenance.

5.

High steam utilization efficiency.

However, the disadvantages are:

1.

Foaming may occur during heating.

2.

Direct contact of the medium with steam requires that allowance be made for condense dilution and requires "clean" steam, free from anticorrosion additives.

In some cases the injection system is combined with flash cooling, where the sterilized medium is cooled by passing it through an expansion valve into a vacuum chamber. Cooling then occurs virtually instantly. A flow chart of a continuous sterilization system using direct steam injection is shown in Fig. 5.13. In some cases a combination of direct and indirect heat exchangers may be used (Svensson, 1988). This is especially true for starch-containing broths when steam injection is used for the preheating step. By raising the temperature virtually instantaneously, the critical gelatinization temperature of the starch is passed through very quickly and the increase in viscosity normally associated with heated starch colloids can be reduced.

Figure 5.13. Flow Diagram of a Typical Continuous Injector-Flash Cooler Sterilizer

The most widely used continuous sterilization system is that based on the spiral heat exchangers and a typical layout is shown in Fig. 5.14. Junker et al. (2006) described the replacement of a pilot-plant direct-steam injection process with one based on indirect spiral heat exchangers, thus avoiding the problems of the medium quality being affected by fluctuations in the plant steam quality—particularly the presence of steam anticorrosion additives. The key features of this plant were:

Figure 5.14. Flow Diagram of a Typical Continuous Sterilization System Employing Spiral Heat Exchangers

Alfa-Laval Engineering Ltd., Camberley, Surrey, UK.

1.

A number of feed tanks from 2,000 to 19,000 dm³ that enabled both a range of medium volumes to be processed and different medium ingredients to be sterilized sequentially and separately.

2.

A recycle tank (also referred to as a circulation or surge tank) and recycle facility so that water could be circulated through the system during cleaning and sterilization of the plant.

3.

A spiral heat exchanger to raise the temperature of the medium to the sterilization temperature.

4.

A "retention loop" to hold the medium at the sterilization temperature for the required time.

5.

A "recuperator" spiral heat exchanger to cool the sterilized medium leaving the retention loop against incoming cold, unsterile medium, thus minimizing heat wastage.

6.

A cooling spiral heat exchanger supplied with chilled water to cool the sterilized medium leaving the recuperator exchanger to the process temperature.

7.

An additional cooling spiral heat exchanger used when the system was used to produce sterile water.

8.

A switching station to direct the product stream from the final coolers:

a.

to a distribution system to the fermenters (if medium is sterile),

b.

to the recycle tank for recirculation through the system during sterilization of the system, and

c.

to the sewer during cleaning of the system or should a malfunction occur.

The Del factor to be achieved in a continuous sterilization process has to be increased with an increase in scale, and this is calculated exactly as described in the consideration of the scale up of batch regimes. Thus, if the volume to be sterilized is increased from 1000 to 10,000 dm³ and the risk of failure is to remain at 1 in 1000 then the Del factor must be increased from 34.5 to 36.8. However, the advantage of the continuous process is that temperature may be used as a variable in scaling up a continuous process so that the increased ▿ factor may be achieved while maintaining the nutrient quality constant. A further advantage of the continuous process is that the distribution system can be devised such that a sterile medium stream can be diverted for use on a small scale. Thus, laboratory or small fermenter scale models of the process can be conducted using medium that has been exposed to exactly the same sterilization regime as the production scale.

Deindoerfer and Humphrey (1961) discussed the application of their Nutrient Quality Criterion (Q) to continuous sterilization scale-up. It will be recalled from our earlier discussion of batch sterilization that:

$Q=A\cdot t\cdot {\text{e}}^{-(E/RT)}.$

Therefore, as for the Del factor equation, by taking natural logarithms, and rearranging, the following equation is obtained

(5.12) $\ln t=\frac{E}{RT}+\ln \frac{Q}{A}.$

Thus, a plot of the natural logarithms of the time required to achieve a certain Q value against the reciprocal of the absolute temperature will yield a straight line, the slope of which is dependent on the activation energy; that is, a very similar plot to Fig. 5.5 for the Del factor relationship. If both plots were superimposed on the same figure, then a continuous sterilization performance chart is obtained. The example put forward by Deindoerfer and Humphrey (1961) is shown in Fig. 5.15. Thus, in Fig. 5.15, each line of a constant Del factor specifies temperature-time regimes giving the same fractional reduction in spore number and each line of a constant nutrient quality criterion specifies temperature–time regimes giving the same destruction of nutrient. By considering the effect of nutrient destruction on product yield, limits may be imposed on Fig. 5.15 indicating the nutrient quality criterion below which no further increase in yield is achieved (ie, the nutrient is in excess) and the nutrient quality criterion at which the product yield is at its lowest (ie, there is no nutrient remaining). Thus, from such a plot a temperature–time regime may be chosen which gives the required Del factor and does not adversely affect the yield of the process.

Figure 5.15. Continuous Sterilization Performance Chart (Deindoerfer & Humphrey, 1961)

The precise adoption of Deindoerfer and Humphrey's approach is possible only if the limiting heat-labile nutrient is identified and the activation energy for its thermal destruction is experimentally determined. As pointed out by Banks (1979), this information may not be available for a complex fermentation medium. However, the compromises proposed by Singh et al. (1989) and Boeck et al. (1989) could allow the application of this approach. As discussed earlier, the Del factor is scale dependent and therefore as the volume to be sterilized is increased so the Del factor should be increased if the probability of achieving sterility is to remain the same. However, the nutrient-quality criterion is not scale dependent so that by changing the temperature–time regime to accommodate the attainment of sterility, the nutrient quality may be adversely affected. Examination of Fig. 5.15 indicates that the only way in which the Del factor may be increased without any change in the nutrient quality criterion is to increase the temperature and to decrease the holding time.

When designing a continuous sterilization process based on spiral heat exchangers, it is important to consider the effect of suspended solids on the sterilization process. Microorganisms contained within solid particles are given considerable protection against the sterilization treatment. If the residence time in the sterilizer is insufficient for heat to penetrate the particle then the fermentation medium may not be rendered sterile. The routine solution to this problem is to "over design" the process and expose the medium to a far more severe regime than may be necessary. Armenante and Li (1993) discussed this problem in considerable detail and produced a model to predict the behavior of a continuous system. Their analysis suggested that the temperature of the particle cores is significantly less than that of the bulk liquid. Furthermore, there is a considerable time lag in heat penetrating to the particle cores, resulting in a very different time–temperature profile for the particles as compared with the liquid medium. Thus, the temperature of the particles may not reach the critical point before they leave the sterilizer and heat penetration into the particles will continue downstream of the sterilizer. Armenante and Li's conclusion is that it is the sterilizer and/or the first cooling exchanger that should be "overdesigned" rather than the length of the holding coil. Remember that the first cooling exchanger transfers a significant amount of heat from the sterile medium to the incoming medium and increasing its surface area would give more opportunity for the heat to penetrate the particles. This, coupled with increasing the temperature or residence time in the sterilizer, would ensure that the particle cores are up to temperature before the holding coil is reached. Also, this work suggests that it is unwise to use the direct steam injection method to heat a particulate medium because, again, there will be insufficient time for the heat to penetrate the particles.

An example of the scale up of sterilization regimes is given by the work of Jain and Buckland (1988) on the production of efrotomycin by Nocardia lactamdurans. In this case, a beneficial interaction appeared to be occurring between the protein nitrogen source and glucose during sterilization, thus making the protein less available but resulting in a more controlled fermentation. When glucose was sterilized separately, the oxygen demand of the subsequent fermentation was excessive and the fermentation terminated prematurely with very poor product formation. On scaling up the fermentation, it was very difficult to attain the correct sterilization conditions using a batch regime. However, continuous sterilization using direct steam injection allowed the design of a precise process producing sterile medium with the required degree of interaction between the ingredients. The identification of this phenomenon was dependent upon careful monitoring of the small-scale fermentation and consideration being given to sterilization as an important scale-up factor.

When a fermentation is scaled up, it is important to appreciate that the inoculum development process is also increased in scale (see Chapter 6) and a larger seed fermenter may have to be employed to generate sufficient inoculum to start the production scale. Thus, the sterilization regime of the seed fermenter (and its medium) will also have to be scaled up. Therefore, the performance of the seed fermentation should be assessed carefully to ensure that the quality of the inoculum is maintained on the larger scale and that it has not been adversely affected by any increase in the severity of the sterilization regime.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780080999531000053

Bioreactor Design Operation

Shijie Liu , in Bioprocess Engineering (Second Edition), 2017

17.7.2.2 Thermal Sterilization in a PFR

We next examine a PFR option for continuous sterilization. Fig. 17.12 shows a diagram of a tubular sterilizer. Viable cell balance in the differential volume between x and x  + dx along the tubular axis gives:

Fig. 17.12. Sterilization in a well-mixed reactor.

(17.39) ${\left.Q{C}_{\mathrm{X}}\right|}_x-{\left.Q{C}_{\mathrm{X}}\right|}_{x+dx}-k_{\mathrm{d}}{C}_{\mathrm{X}}{A}_{\mathrm{c}\mathrm{r}}dx=\frac{\partial \left(C_{\mathrm{X}}{A}_{\mathrm{c}\mathrm{r}}dx\right)}{\partial t}=0$

where x is the tube length coordinate (from the feed point) and A _cr is the cross-sectional area of the tube. Divide Eq. (17.39) by dx and letting dx  →   0, we obtain:

(17.40) $k_{\mathrm{d}}{C}_{\mathrm{X}}{A}_{\mathrm{c}\mathrm{r}}+\frac{d\left(Q{C}_{\mathrm{X}}\right)}{dx}=0$

Integrating Eq. (17.40), we obtain the probability of a contaminant cell surviving the heat treatment:

(17.41) $p\left(\tau \right)=\frac{C_{\mathrm{X}}}{C_{\mathrm{X}0}}={\mathrm{e}}^{-k_{\mathrm{d}}\tau }$

where τ is the spacetime of the fluid medium through the whole length of the sterilization section L, that is:

(17.42) $\tau =\frac{A_{\mathrm{c}\mathrm{r}}L}{Q}$

Therefore, sterilization in a PFR is quite similar to that of the batch process. The probability of the entire population in one reactor-full process fluid (of volume V) not vanishing from the heat treatment is:

(17.43) $P_1\left(\tau \right)=1-P_0\left(\tau \right)=1-{\left[1-p\left(\tau \right)\right]}^{C_{\mathrm{X}0}V}=1-{\left(1-{\mathrm{e}}^{-k_{\mathrm{d}}\tau }\right)}^{C_{\mathrm{X}0}V}\approx \frac{C_{\mathrm{X}0}V{\mathrm{e}}^{-k_{\mathrm{d}}\tau }}{1+\frac{C_{\mathrm{X}0}V-1}{2}{\mathrm{e}}^{-k_{\mathrm{d}}\tau }}$

which corresponds to the dimensionless sterilization time of:

(17.44) $t_{\mathrm{S}}=k_{\mathrm{d}}\tau -ln\left(C_{\mathrm{X}0}V\right)$

Eq. (17.44) is nearly identical to Eq. (17.25). Thus, sterilization in a PFR is just like sterilization in a batch reactor. However, the heating up and cooling down requirements are not as stringent as batch sterilization.

Continuous sterilization in a PFR set-up, particularly a high-temperature, short-exposure time process, can significantly reduce damage to medium ingredients while achieving high levels of cell destruction. Other advantages include improved steam economy and more reliable scale-up. The amount of steam needed for continuous tube-flow sterilization is 20–25% that used in batch processes. The time required is also significantly reduced because heating and cooling are virtually instantaneous. Therefore, for sterilizing process fluids, it is advantageous to design the sterilization in a PFR setting.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780444637833000174

Thermal processes, methods and equipment

Zeki Berk , in Food Process Engineering and Technology, 2009

18.3.2 Bulk heating – holding – bulk cooling – cold filling – sealing

In this mode of operation, the entire process is carried in a system consisting of heat exchangers and piping. The product leaves the system after continuous sterilization or pasteurization (depending on the time–temperature profile), holding and continuous cooling. If filling and sealing occur in open space, recontamination of the cold product may occur. This may not be objectionable if the product is to be marketed under refrigeration and its planned marketing time is sufficiently short. This is the case of pasteurized milk and dairy products (Walstra et al., 2005). A typical process for the continuous pasteurization of liquid milk is schematically shown in Figure 18.15.

Figure 18.15. Schema of the continuous flash pasteurization of milk

The pasteurizer, in this case, is a plate heat exchanger consisting of three sections. Raw milk is admitted to Section 1 (preheating, regeneration) where it is heated to about 55°C, by hot pasteurized milk flowing on the other side of the plates. It then passes to Section 2 (heating) where it is heated to the specified process temperature, with hot water on the other side. The hot water for this purpose is heated by steam and circulated in close circuit by appropriate pumps (not shown in the Figure). The milk, thus pasteurized, passes through a holding tube of appropriate dimensions (according to the specified holding time). A three-way control valve (flow diversion valve, FDV), governed by the control system, diverts any under-processed portion back to the feed tank. Milk that has reached the specified process temperature (normally, minimum 72°C) is sent to Section 1 where its is partially cooled by the incoming raw milk and then to Section 3 (cooling) for final cooling with cooling water and ice-water and, finally, to packaging as pasteurized and cooled (normally 4°C ) milk.

Although considerable energy saving is achieved by including in the system the 'regeneration' section, the two-stage heating has additional technological benefits. The intermediate temperature of about 55°C is ideal for centrifugal separation of fat and for high pressure homogenization. These processes are then performed on the preheated milk from the regeneration section.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780123736604000181

Principles for the Application of Nanomaterials in Environmental Pollution Control and Resource Reutilization

Liming Yang , ... Sheng-Lian Luo , in Nanomaterials for the Removal of Pollutants and Resource Reutilization, 2019

1.2.3.4 Disinfection

Disinfection is the last but critical process in wastewater treatment and water supply. Providing effective disinfection whilst minimizing the formation of toxic disinfection by-products (DBPs) poses a great challenge. An ideal disinfectant should possess the following properties: (1) broad antimicrobial spectrum, (2) fewer harmful byproducts and low toxicity, and (3) easy operation. Conventional disinfectants (such as liquid chlorine, chlorine dioxide, and ozone) readily form DBPs, such as halogenated compounds, nitrosamines, and bromate. In contrast, ultraviolet (UV) ray disinfection produces minimal DBPs; however, it requires water pretreatment due to limited light transmission and continuous sterilization ability cannot be retained. Therefore it is necessary to develop alternative methods that could enhance the robustness of disinfection while avoiding or reducing the formation of DBPs.

Nanotechnology has the ability to develop some nanomaterials to fight against antimicrobial-resistant bacteria. Several functional nanomaterials have been equipped with unique antimicrobial properties, including nano-Ag [19], chitosan nanoparticles, nano-ZnO [20], nano-TiO₂ [21], nano-Ce₂O₄, and carbon-based nanomaterials [22]. Compared with bulk materials, such small size particles with a high surface-to-volume ratio could closely interact with microorganisms. They trigger different biological responses, and hence exhibit promising antimicrobial performance (the ability to reduce the attachment and viability of microbes). In detail, they kill microbial cells by releasing toxic metal ions (e.g., Zn^{2   +} and Ag⁺), damaging cell membrane structure through direct contact (e.g., chitosan) or generating reactive oxygen species through photocatalysis on TiO₂ semiconductor nanoparticles. Unlike conventional chemical disinfectants, these antimicrobial nanomaterials inactivated microorganisms through a more sustainable approach without strong oxidation and thus have a lower tendency to form DBPs [23, 24].

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128148372000019

Industrial Biotechnology and Commodity Products

L.E. Erickson , in Comprehensive Biotechnology (Second Edition), 2011

3.55.2 Classifications of Bioreactors

There are several common methods to classify bioreactors. Batch, fed batch, and continuous operation are all commonly found in practice. Ethanol production is often a batch operation, while anaerobic digestion is usually a continuous process. Batch fermentations may be continuous with respect to the gas phase. There may be some effort to retain some or all of the solids in a continuously operating anaerobic digester.

The mixing in a bioreactor may be used for classification. In most bioreactors, the liquid phase is modeled as a completely mixed tank; however, there are some flow applications in which axial mixing is intermediate between the extremes of complete mixing and plug flow. The continuous sterilization of milk and other fluids is an example in which the desired flow model is that of plug flow; however, it is not possible to eliminate all axial mixing. The dispersion model can be used to represent flow systems with axial mixing, such as flow in extruders and jet cookers, both of which are used for processing starch products [7, 22, 25, 27].

In many applications, a single strain is present that has desired properties and abilities. These pure culture processes allow applications of molecular biotechnology to be used to produce products of commercial value. Biological wastewater treatment and anaerobic digestion are examples of mixed culture processes, in which there is positive value of having a mixed culture present. When there are a variety of substrates, there is a need for a mixed culture to digest them. Yogurt fermentations are carried out with mixed cultures of Streptococcus thermophilus and Lactobacillus bulgaricus that ferment milk very effectively [12].

Classification with respect to oxygen includes aerobic, microaerobic, and anaerobic processes. Oxygen is often supplied by air sparging into the bioreactor; however, in animal cell cultures, diffusion of oxygen through tubes or membranes may be employed to reduce turbulence and cell damage. Nitrate or sulfate may be used as source of oxygen in some applications.

Classification may be with respect to the nature of the process. The production of yeast for baking is an example where the microorganisms are the product of interest. The ethanol fermentation is an example of a fermentation with a simple product, while the penicillin fermentation is an example of a more complex product. Wastewater treatment is an example of a process in which biodegradation is the desired service. Those processes in which enzymes are used in a bioreactor to produce a product or provide a service are classified as enzymatic processes. When the enzyme is used in diagnostic applications, the bioreactor may be a simple small blob on a kit that is used to look for color change or an electrical signal after the substrate is applied [26]. There are many products made by biocatalysis that involve multiple steps, in which enzymes are employed in part of the process, while chemical synthesis is used in other parts of the process [6, 10, 16, 18]. Food processing applications include baking of bread, production of yogurt, cheese manufacture, and conversion of cucumbers to pickles. In the remediation of contaminated soil and groundwater, biodegradation may occur in a soil/water system that is designed to make use of native microorganisms or in some cases an added culture that has the desired biodegradation capacities [20].

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780080885049002361

Flow of Multi-phase Mixtures in Pipes

R.P. Chhabra , J.F. Richardson , in Non-Newtonian Flow and Applied Rheology, 2008

4.3.2 RTD and Slip Velocity

The heat treatment to which a food particle is subjected is directly influenced by its residence (or contact) time in the plant, especially in the holding section. Therefore, reliable knowledge of RTD is very important, although, the RTD of single phase fluids in continuous sterilization or pasteurization in tube flows is reasonably well understood. For instance, for the laminar flow of Newtonian fluids in a tube, the centerline velocity is twice of the mean velocity. Therefore, the fluid elements in the center of the tube would have the shortest residence time. For highly shear-thinning power-law fluids, the velocity profile is flattened and the ratio of the mean velocity to the centerline velocity is given by equation (3.11), namely, {(n+1)/(3n+1)}. The greater is the degree of shear-thinning (small value of power-law index n), flatter is the velocity profile across the tube cross-section and clearly, this is advantageous in reducing the width of the RTD. Based on experimental results, the estimation of the required length of holding tube is typically based on a maximum (centerline) velocity of twice the mean velocity (as in laminar flow of a Newtonian fluid) to ensure a safe process for all conditions (Lareo et al., 1997b).

On the other hand, however, the corresponding situation for the flow of food suspensions is less clear. Most of the advances made in this field are based on experimental observations aided by empirical and/or phenomenological considerations. Model studies with single food (real and prototype) particles suspended in Newtonian and in shear-thinning vehicles in tube flows have provided useful insights. For instance, Liu et al. (1993) investigated the behaviour of 6   mm carrot cubes in water (up to ∼35% concentration) in a 44   mm diameter horizontal pipe. Depending upon the mean mixture velocity, they identified six possible flow regimes (with increasing flow rate): stationary particles at the wall, stick-slide type of flow with a variable velocity, sliding along the wall, turning and tumbling but still a great majority of particles in contact with the wall, saltation and suspended. Using dimensional analysis, Liu et al. (1993) correlated the ratio of the particle to fluid velocity for single particles with the reciprocal of the particle Froude number, defined as $[V_{\mathrm{m}}/\sqrt{\mathit{gd}(s-1)}]$ . For concentrated systems, the particle velocity was lower than that of the single particle, and the data on relative velocity still correlated well with the inverse of the Froude number, but with a significant scatter. One can then infer the mean residence time for particles from a knowledge of the slip velocity.

Several other studies on the RTD of food-like particles in shear-thinning and viscoplastic carriers in horizontal tubes have been reviewed by Lareo et al. (1997b), but none of these is conclusive. For instance, the mean and minimum residence times and the standard deviation of their RTDs of particles all show a complex dependence on the particle Reynolds number, power-law behaviour index, particle-to-tube diameter ratio and particle concentration (Sandeep and Zuritz, 1994, 1995). Broadly speaking, the velocity profile flattens with the increasing concentration of solids and, as the particles are nearly neutrally buoyant, most of them travel with an approximate uniform velocity. However, the particles never seem to achieve the maximum expected velocity (based on the assumption of power-law fluid behaviour) even at the axis of the tube (Tucker and Withers, 1994). Under certain conditions, especially at low solid loadings, it is observed that the particle RTDs show two distinct peaks thereby suggesting the formation of two groups of particles moving at different velocities, i.e., segregation of slow moving particles near the wall and of fast moving particles in the middle of the tube. Naturally, in this flow regime, particles near the tube walls will have much longer residence time than their counterparts near the axis of the tube thereby resulting in uneven quality of product.

Similarly, several commercial food processes (i.e. the APV baker ohmic heater) use vertical or inclined pipe work through which solid–liquid food mixtures are passed. Some attempts have been made to predict the RTD of solids and/or their slip velocities in vertical flow. For instance, Lareo et al. (1997a), Lareo and Fryer (1998) and Fairhurst et al. (1999) have investigated the RTD for real and model food particles in shear-thinning polymer solutions in vertical (upward and downward) flows. Lareo et al. (1997a) reported significant migration of particles across streamlines at volume concentrations up to about 10% using 6–10   mm carrot particles in a 44   mm diameter pipe. Over a range of velocities and particle loadings, they observed a plug-like flow wherein the particles segregated in the region ∼0.2≤(r/R)≤∼0.5, and were seen to be moving with the approximately same axial velocity. Also, a great majority of the particles were seen to be travelling faster than the mean mixture velocity (presumably due to the segregation in the middle of the tube). Therefore, the mean residence time of the particles was less than that of the mixture, though not significantly so. The RTD of particles also appeared to be strongly influenced by whether the liquid was simply shear thinning or also exhibited a yield stress. On the other hand, the minimum and mean residence (or passage) times showed a positive correlation with the concentration in the discharged mixture.

Similarly, preliminary studies have also been reported on RTDs in scraped surface heat exchanger (SSHE) and in helical coils with a view to achieve long contact time, albeit results are strongly dependent on the detailed geometrical design of the equipment (Tucker and Heydon, 1998; Sandeep et al., 2000; Cheng et al., 2005). Although both these geometries offer longer contact times and space economy than a straight tube, but only at the expense of added complexity of the flow.

Thus, in summary, although the advantages of continuous processing of foodstuffs are well recognized, due to the inadequate understanding of solid–liquid flows even in straight tubes, very little definitive information is yet available on sound process design and performance of the equipment used in the thermal treatment of foodstuffs.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780750685320000044

Engineering Fundamentals of Biotechnology

D. Pollard , in Comprehensive Biotechnology (Second Edition), 2011

2.68.1 Introduction

The aseptic operations during any bioprocess are vital for controlling the desired level of microbial load (bioburden) of a given biological product. Manufacturers follow the standards and procedures set by the regulatory agencies [1, 2] to ensure product quality. This minimizes the risk to patients of exposure to unacceptable levels of contamination such as failure to maintain product sterility. Practical and experience-based practices, not captured in regulatory documents, have been defined by organizations [3–5] and harmonization conferences [6, 7]. These regulatory compliant procedures are based upon rational, evidence-based science, and engineering with incorporation of risk assessment analysis.

The manufacturing of sterile products is acknowledged to be the most difficult of all pharmaceutical production activities [3]. For bioprocess production, such as aqueous protein or monoclonal antibody solutions ( Figure 1 ), some form of aseptic operations usually encompass every production step from fermentation, purification, formulation, and fill. It is a regulatory requirement to assure that culture purity (single organism) is maintained from the master cell bank and throughout the upstream step ( Figure 1 ). During purification, maintaining the low bioburden specifications of the drug substance is completed using a combination of chemical sanitization of equipment and filtration of all buffers, including those used for formulation. After formulation the final drug product is filter sterilized and aseptically filled into the final container (vial, syringe, or IV bag) and, in some cases, lyophilized.

Figure 1. Bioprocess map for production of a therapeutic protein or monoclonal antibody using microbial fermentation using a Pichia expression host. Stages for sterility and bioburden testing are indicated. Steps that have disposable options are indicated.

Managing contamination to minimal levels covers a wide range of activities, including facility design, equipment setup, process operations, process validation, process monitoring, and personnel training ( Figure 2 ). Despite all these efforts including attention to detail for process operations, contaminations will always occur as no microbial fermentation or cell culture facility is contamination free. So procedures need to be in place for vigorous investigations of contamination to build experience and a knowledge database. Literature contamination rates of 5–30% have been described for microbial fermentation with a contamination probability of 1 out of 100 acceptable [8]. Examples of rates <1% are considered commendable and indicated as good performance [8–10], while rates of 2% have been recorded for animal cell culture [8, 11, 12]. Facility improvements have shown to lower contamination rates. For example, the contamination rate for monosodium glutamate production was reduced from 4.5% to 0.6% by a combination of sparger air system upgrades, installation of a laminar flow hood in the inoculum room, and repair of holes in the heat exchangers of the continuous sterilization system [11].

Figure 2. Integrated approach for control of contamination in bioprocessing.

For the microbial fermentation or cell culture process, microbial contamination is the most common cause of process failure over mechanical, electrical, or instrumentation problems that occur [10]. Microbial contamination can impact the process by changing the chemical conditions such as the conversion of nutrients to unwanted impurities, changing the pH, and triggering the formation of enzymes leading to product degradation [8]. Historically, there are only a few examples of facilities that will continue to process contaminated batches beyond fermentation. These were usually for small-molecule natural products/anti-infective products where the subsequent chemical steps achieved sufficient purity and removed contaminants. This is certainly not an option for the injectable products for biologics and vaccines. Upon contamination detection, the entire batch is discarded, equipment shutdown, and a failure investigation initiated. This causes substantial losses of time, materials, and revenue, with disruption to the facility schedule.

Contaminants vary by product type but the most frequent microbial contaminants are from two forms: (1) fast-growing spore-forming Gram-positive bacteria such as Bacillus subtilis, associated with incomplete sterilization such as from large-medium particles or residual dried batch in vessel crevices [10, 13]; and (2) Gram-negative rods, which are indicative of cooling water leak [14], water in the inlet air, or incomplete filter sterilization. Gram-positive bacteria often enter from non-sterile air [8], owing to improper air filter installation, sterilization, or integrity [15]. Multiple contaminants are usually indicative of general sterilization failure [16]. Mycoplasma is an important contamination to monitor for cell culture processes [17–19]. Mycoplasmas lack a cell wall, have filterability at 0.22   μm, and are easily killed at 60   °C. Cell culture media components are heat-labile sensitive, so sterilization by filtration is the only option. Mycoplasma infections can overwhelm production cell cultures achieving high densities (10⁶–10⁷ colony-forming unit (CFU) ml⁻¹) but visually no turbidity is observed [19]. Twenty species of mycoplasmas are known to cause cell culture issues and five have shown to give >95% of contaminations (Mycoplasma arginini, M. fermentas, M. hyorhinis, M. orale, and Acholeplasma laidlawii) [19] . Contamination sources are commonly from human operators or from the cell lines. Viral contamination, via endogenous viruses or adventitious viral agents, is an important concern for cell culture [17, 20]. Cell lines contain retrovirus-coding sequences in their genomes and therefore inherently express retrovirus particles during production. Adventitious viral agents may be introduced through the use of cell lines derived from infected animals or virus-contaminated reagents or serum components. Safety assurance is accomplished by the combination of raw material control/testing, master cell and working bank testing, in-process control testing, and virus clearance studies. Typical viral testing includes a panel of viruses, ranging in size from 17 to 400   nm, such as bovine viruses (viral diarrhea virus, adenovirus, polyoma virus), reovirus, cache virus, and murine minute virus. Reported viral infections of recombinant CHO cell lines include murine minute virus, a parvovirus [21], and epizootic hemorrhagic disease virus [22].

The risk of contamination has to be evaluated for each particular bioprocess. Subtle changes in operating conditions between processes can have a large impact on the susceptibility to contamination. Certain factors have been identified that lower the risk of contamination such as pH range (<5 and >8), low initial bioburden of the media before sterilization, high osmotic pressure, high or low carbon concentrations, switching from complex media with insoluble solids content to soluble defined media [16, 23], or applying temperatures above 60   °C.

Minimizing the risk of microbial contamination is a combination of prevention activities and contamination monitoring ( Figure 2 ). Contamination risks occur in aging facilities that are susceptible to mechanical failures, and also new facilities that have operational unknowns [16]. A balance needs to be addressed between increasing time for preventative maintenance (PM) to reduce failures versus fast turnaround times to maximize the productivity of the facility [11]. Process design, testing, and training are all important. Protocols must be in place for each new bioprocess to minimize and investigate microbial contaminations [10, 16]. This is influenced by the nature of the fermentation/contaminant, equipment design, process operation procedures, and the microbiological process controls implemented ( Figure 2 ).

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780080885049001483

rivastiver1982.blogspot.com

Source: https://www.sciencedirect.com/topics/engineering/continuous-sterilisation

Share this post