Fig. 7.8. Joint seals for glass-glass, glass-metal and metal-metal; (a) gasket; (b) lip seal; (c) 'O' ring in groove.
ment levels comparable with the U.K. B3 (Titchener-Hooker et al., 1993).
Titchener-Hooker et al. (1993) criticized Chapman's proposal to use two seals without a steam trace for a number of reasons including:
a. Double seals are more difficult to assemble correctly.
b. It is difficult to detect failure of one seal of a pair during operation or assembly.
c. Neither of the two seals can be tested independently.
d. Dead spaces between two seals must be considered to be contaminated.
Leaver (1994) and Titchener-Hooker et al. (1993) consider that correctly fitted single static seals can provide adequate containment for most processes and double static seals with a steam trace should be strictly limited to the small number of processes for which an extreme level of protection may be required.
Normally in the design and construction of a fermenter there must be adequate provision for temperature control which will affect the design of the vessel body. Heat will be produced by microbial activity and mechanical agitation and if the heat generated by these two processes is not ideal for the particular manufacturing process then heat may have to be added to, or removed from, the system. On a laboratory scale little heat is normally generated and extra heat has to be provided by placing the fermenter in a thermostatically controlled bath, or by the use of internal heating coils or a heating jacket through which water is circulated or by a silicone heating jacket. The silicone jacket consists of a double silicone rubber mat with heating wires between the two mats; it is wrapped around the vessel and held in place by Velcro strips (Applikon, 1989).
Once a certain size has been exceeded, the surface area covered by the jacket becomes too small to remove the heat produced by the fermentation. When this situation occurs internal coils must be used and cold water is circulated to achieve the correct temperature (Jackson, 1990). Different types of fermentation will influence the maximum size of vessel that can be used with jackets alone.
It is impossible to specify accurately the necessary cooling surface of a fermenter since the temperature of the cooling water, the sterilization process, the cultiva-
don temperature, the type of micro-organism and the energy supplied by stirring can vary considerably in different processes. A cooling area of 50 to 70 m2 may be taken as average for a 55,000 dm3 fermenter and with a coolant temperature of 14° the fermenter may be cooled from 120° to 30° in 2.5 to 4 hours without stirring. The consumption of cooling water in this size of vessel during a bacterial fermentation ranges from 500 to 2000 dm3 h^1, while fungi might need 2000 to 10,000 dm3 h 1 (Muller and Kieslich, 1966), due to the lower optimum temperature for growth.
To make an accurate estimate of heating/cooling requirements for a specific process it is important to consider the contributing factors. An overal energy balance for a fermenter during normal operation can be written as:
= heat generation rate due to microbial metabolism, = heat generation rate due to mechanical agitation, = heat generation rate due to aeration power imput, = heat accumulation rate by the system, = heat transfer rate to the surroundings and/or heat exchanger, = heat loss rate by evaporation, = rate of sensible enthalpy gain by the flow streams (exit — inlet).
This equation can be rearranged as:
öexch = Ö met + Öag + ögas - Ö acc Ösen Öevap
öexch is die heat which will have to be removed by a cooling system.
Atkinson and Mavituna (1991b) have presented data to estimate <2mct for a range of substrates, methods to calculate power input for gag and ßgas, a formula to calculate the sensible heat loss for flow streams (ßsen) and a method to calculate the heat loss due to evaporation (öevap)- Cooney et al. (1969) determined representative low and high heats of fermentation values for Bacillus subtilis grown on molasses at 37° (Table 7. 6). They concluded that ßevap and Qscn are small contributory factors and Qacc = 0 in a steady-state system. <2cvap can also be eliminated by using a saturated air stream at the temperature of the broth.
Was this article helpful?