Relaciones Interpersonales

Mass balances were used to study the significance of different levels of dewatering around the primary separation stage for the recovery of intracellular protein from a fermentation of baker's yeast, as shown in Figure 2.1. A sample spreadsheet is shown in Appendix 1 (a).

2.1.1 Format

The volume basis used for examining interaction of process variables in the dewatering spreadsheet was the process analysis standard of 1 L.

Figure 2.1.1: Mass balance flo w sheet. Boxed section indicates operations considered in dewatering mass balance.

□

F e rm e n ta tio n H arvesting D isruption Floccu la tio n D e w a t e r i n g C la rific a tio n Filtra tion C h ro m a to g ra p h ic p u r ific a tio n

Yeast fermentations typically produce 60 wwg L'^ whole cells (Dehgani 1996). Cells are concentrated during the harvesting process by a factor of up to

6x (10 X is possible according to Aronsson 1987) in this spreadsheet. This factor

was varied during experiments with the spreadsheet in order to examine the effect of feed concentration on dewatering.

Yeast cell composition was obtained from selected texts (Mateles and Tannenbaum 1968, Harrison 1967, Moo Young and Gregory 1986) and could be altered for a different microorganism, or for a recombinant strain with increased DNA content (rDNA process). Intracellular water occupied 70% of the total volume (Ward 1989). Cell components shown fed to the homogeniser were included in the total sum only as components of the whole cells. Concentrations were calculated throughout as fractions of the total mass.

The harvested process stream was disrupted in an industrial homogeniser and cell contents release calculated according to Hetherington et al (1971) for protein release. Protein release is linearly proportional to cell concentration not exceeding 60% by volume (660 g L'^). It was assumed throughout that the process temperature would be controlled in a way to minimise product degradation due to protease action. Neither temperature nor product degradation were not variable parameters in the spreadsheet mass balance.

A flocculation step using either borax or PEI extracted cell debris and the major soluble contaminants into a solid phase of the process stream according to data from Milbum et al (1990). The process stream was divided into liquid and solid phases at this stage, where the composition of the liquid product phase was calculated first. It was assumed that no cells would remain in solution, no protein product would be precipitated, 50% carbohydrates would remain in solution and 4.4% of the extracellular water was chemically bound to the flocculated material (Wiesmann and Binder 1982). The amount of PEI required was found from a stoichiometric assessment of yeast cell debris flocculation (Milbum et al 1990). Flocculant was added at 10% of the process stream volume according to standard industrial practice.

The physical separation of the solid and liquid phases was said to occur in a single unit operation where there was no carryover of contaminating solids into the product stream. In this ideal separation step, the only product loss was through liquor entrained in the solids discharge.

Having determined that no solids were present in the product stream, then all were present in the sediment discharge. The solids dryness specification determined the total mass of solids discharged and hence the mass of entrained liquor. Any liquor not entrained remained in the product stream.

The step protein yield for this and other unit operations was calculated as the fi-action of protein dissolved in the product stream relative to the total present in the feed.

Experiments were carried out with the spreadsheet where the feed stream cell concentration was varied fi*om a typical yeast fermentation broth ( 2 0 gdcw

L'^) to cells harvested using a centrifuge (200 gdcw L'*). This was repeated for a range of sediment solids concentrations from 20-60 % dwt/wwt, given as kg solids / kg total mass.

2.1.2 Results

Figure (2.1.2) shows different levels of dewatering as a fimction of the input solids concentration and the overall yield of the operation. Taking typical values as an input concentration of 1 0 0 gdcw L'^ and a step yield in excess of

90%, then it is clear that this can only be achieved if a level of dewatering greater than 60% (expressed as 0 . 6 kg dry solids per kg total sediment) is

realised. Current dewatering technology using the scroll decanter centrifuge is limited to 0.3 kgkg'\

Overall recovery may be enhanced by dilution of the process stream at the flocculation stage to avoid entrainment of concentrated product in poorly dewatered solids. However, dilution may not be a feasible option for rDNA processes where the cost of containing large volumes of material is prohibitive. This study will aim to investigate how dewatering can be maximised in a typically concentrated process system of homogenised yeast flocculated with PEI.

100 0.6 kg /k g 0.4 kg /k g (U o a o I o C /D Lw O <v 0.2 kg/kg 4 0 5 0 0 4 0 0 6 0 0 3 0 0 1 0 0 200 0

Initial cell con cen tratio n (g L '')

F ig u r e 2.1.2: S im ulation o f the effect o f d e w a te rin g on the rec o v e ry o f in tra ce llu la r yeast protein. S edim ent d e w a te rin g given in kg dry se d im e n t per kg total sedim ent.