Plant freeze-drying

At shelf temperatures of 100 °C, approx. 2000— fOOO kJ/h m2 are transmitted, depending on the product temperature. At lower shelf temperatures, as is usual in freeze-drying plants for pharmaceutical products, q values between 500 and 1500 kJ/m2 can be expected. However for e = 0.12, these data are reduced by a factor of 0.12. 

If the shelf and the tray are as planar as technically possible, the plot marked s = 0 applies. At 0.2 mbar, a heat transfer coefficient of approx. 85 kJ/m2 h °C can be achieved, rising by a factor of two at 1 mbar. In a well designed freeze drying plant with planar trays or vials a heat transfer coefficient of 160 kJ/h m2 °C at 0.9 mbar is possible (Fig. 1.59), while at a pressure of 0.45 mbar, approx. 120 kJ/h m2 °C (Table 1.9) is measured for the heat transfer coefficient A"tot. To sublimate 1 kg of ice per hour and m2 with a coefficient of... 

Fig. 1.65. Schema of the freeze drying plant in which the tests shown in Fig. 1.62 have been carried out.

To insure an undisturbed water vapor transport (see Section 1.2.4) the leak rate of a freeze drying plant must allows BTM with sufficient accuracy. This applies for vapor pressure in the ice temperature range between -45 °C to -10 °C, corresponding to 0.07 mbar to 2.5 mbar. 

With this leak rate will be in the range of 10 3 mbar L/s. If the leak rate of a plant is stable and known, it can be accounted for in the DR-value. In a normal operation one would expect that a 100-L chamber is loaded with 2.5 kg of liquid product, containing 250 g solids, the leak rate could than be four fold larger, as mentioned above. 

The water vapor transport in a freeze drying plant can be described schematically with the aid of Fig. 1.86 The ice (1) is transformed into vapor and has to flow out of the container (2) into the chamber (4). Between the chamber wall or any other limitation an area (3, FI) is necessary. The vapor flows then through the area (F2) into the condenser (7), having surface of (F3) on which the water vapor will mostly condenses. A mixture of remaining water vapor and permanent gas is pumped through (8), (9) and (10) by a vacuum pump (11). 

Fig. 1.86. Schema for the estimation of the water vapor transport in a freeze drying plant.

Fig. 2.8.1. Freeze drying plant for flasks and bottles, 16 connections 1/2", titan condenser for maximum 3 kg of ice in 24 h, condenser temperature -55 °C or -85 °C (Flexi-Dry MP, FTS Systems, Inc., Stone Ridge, New York).

Fig. 2.8.2. Freeze drying plant for flasks or bottles, 35 connections NS 29/32, maximum 7.5 kg ice in 24 h, Tm down to -53 °C (Lyowall Firma AMSCO Finn-Aqua GmbH, D-50354 Hiirth).

Fig. 2.10. Freeze drying plant of the type in Fig. 2.9 (a). 1600 cm2 temperature-controlled shelf area stoppering device for vials on four shelves, valve between chamber and condenser, for BTM and DR-measure-ments, freezing is possible between the condenser coils or in the shelves if they are cooled and heated by brine from a thermostat, Tco down to -55 °C (LYOVAC GT 2, AMSCO Finn-Aqua, D-50354 Hurth). 

Fig. 2.13.2. Freeze drying plant, in which the trays are placed on to slide rails in between two heating plates by a lift and moved in position by pushing the trays by the last one from the lift (System CONRAD , ATLAS INDUSTRIES A/S, DK-2750 Ballerup).

Fig. 2.19. Plots of a test to determine the specific water vapor flow or the water vapor speed in a production freeze drying plant with approx. 30 m2 shelf area. For the tests 300 kg of distilled water were filled into ribbed trays, which were placed on the shelves. Six RTD were placed in different trays and frozen with the water. The RTD temperatures and BTM measurements were practically identical, because the RTD were always immersed in the ice, and during the test only 25 % of the ice was sublimated.

Fig. 2.21. Temperature of the shelf in a freeze drying plant as a function of cooling time, calculated for four different refrigerants.

Fig. 2.22. Cycle cooled by LN2 of a condenser in a freeze drying plant. A part of the N2 is cooled in a heat exchanger and pumped back in the cycle by a jet pump.

The vacuum pumping system in a freeze drying plant has to fulfil two tasks ... 

Fig. 2.28. Schema of a vacuum pump set, designed for a production freeze drying plant.

In the first freeze drying plant [2.9] phosphorous pentoxide has been used to absorb the water vapor, but to day this technique is outdated. The development of refrigeration technology with compressors or NH3 absorption and the availability of LN2 has replaced water absorption by silica gel or similar products. Some trials aiming to revive this technology are detailed in Section 1.2.6. 

The leak rates of a freeze drying plant can be measured at the empty plant with the condenser cooled and the shelves heated by measuring the pressure rise per time multiplied by the installation volume in the dimension (mbar L/s). It should be noted, that the plant has to be evacuated for several hours, e. g. down to 10-2 mbar, before the pressure rise measurements, to avoid the influence of small amounts of ice and the desorption of gas from the surfaces. Furthermore, the pressure rise should be measured up to 0.2 or 0.4 mbar to detect possible gas desorption. Only if the pressure rise has been for some time proportional with time (Fig. 2.33.1), it represents a leak rate, which is defined as... 

---