4.4.1 Continuous Fermentations for Ethanol Production

With enormous strides made in the development of high-alcohol fermentations pre­pared using high — and very-high-gravity media, and compatible with low operating temperatures, fast turnaround times, and high conversion efficiency from carbohy­drate feedstocks, what has the “traditional” ethanol industry achieved with fermen — tor hardware and process design and control?

Beginning in the late 1950s in New Zealand, the Dominion Breweries intro­duced and patented a novel “continuous” brewing process in which part of the fer­mented beer wort was recycled back to the wort of the start of the fermentation step (figure 4.8).149 Within two years, a rival continuous process had been announced (and patented) in Canada by Labatt Breweries, and independently conceived research was being published by the Brewing Industry Research Foundation in the United King — dom.150 By 1966, the concept had been simplified and reduced to a single-tower structure in which the bulk of the yeast cells were retained (for up to 400 hours) while wort might only reside in the highly anaerobic conditions in most of the tower for only 4 hours before a “beer” product emerged.[36] 151

However, despite much initial enthusiasm and the evolution of technologies into a family tree of “open” (where yeast cells emerge at rapid rate from the process), “closed” (where yeast cells are mostly retained), “homogenous” (approximating a stan­dard stirred-tank fermentor), and “heterogeneous” (with gradients of cells, substrates, and products across several vessels) systems devised for the continuous brewing of beer, very few were developed to the production scale and most became defunct. The innovations failed the test of four basic parameters of brewing practice:150

1. Commercial brewers offer multiple product lines and flexible production schedules to a marketplace that has become more sophisticated, discern­ing, and advertisement-influenced.

2. Continuous systems offer a fixed rate of beer production (and increasing flow rate tends to lead to wash out).

3. Only highly flocculent strains of yeast can be used.

4. Desired flavors and aromas cannot always be generated, whereas undesir­able levels of malodorous compounds are easily formed.

Moreover, there were contemporaneous developments in brewing technology that led, for example, to the marked reduction in malting and fermentation time in batch processes, the losses during malting and wort boiling, increased a-acid content and hops (and, consequently, reductions in hop usage), and the elimination of the prolonged “lagering” storage necessary for lager beer production (an increasing popular product line world­wide). Such incremental improvements were easy to retrofit in established (often vener­ably old) brewing facilities, which demanded mostly tried and trusted hardware for new sites. Other than in New Zealand, continuous brewing systems have proved generally unpopular — and the rise of globally recognized brands has resulted in the backlash of movements for traditional, “real ale,” locally produced (in microbreweries) options, continuing the trend of consumer choice against a perceived blandness in the output of the market leaders. Large multinationals, offering a branded product for global markets, remain the most likely to invest in “high-technology” brewing.

In contrast, industrial and fuel ethanol production is immune to such drivers, and by the late 1970s, the cold-shouldering of new technologies by potable alcohol producers had persuaded academic and research groupings of chemical engineers to focus on eth­anol as an alternative fuel with a renewed interest in highly engineered solutions to the demands for process intensification, process control, and cost reduction. For example, [37]

vessel, a high-productivity process was devised that incorporated sparging with O2 to support active cell growth over a prolonged period and bleed out the fermented broth to withdraw nonvolatile compounds of potential toxicity to the producing cells; for a 95% ethanol product, such technology offered a 50% reduction in production costs over batch fermentation.152

• A single-stage gas-lift tower fermentor with a highly flocculent yeast was designed to run with nearly total retention of the cell population and gener­ate a clear liquid effluent; analysis of the vapor-phase ethanol concentration in the headspace gas gave, via computer control, an accurate control of input and output flow rates.153

• A more straightforward use of laboratory-type continuous fermentors oper­ating on the fluid overflow principle and not requiring a flocculent yeast achieved a maximal ethanol yield (89% of the theoretical from carbohy­drate) at a low dilution rate (0.05/hr), showed 95% utilization of the inflow­ing sugars up to a dilution rate of 0.15/hr, and only suffered from washout at 0.41/hr; the fermentation was operated with a solubilized mixed substrate of sugars from Jerusalem artichoke, a plant species capable of a very high carbohydrate yield on poor soils with little fertilizer application.154

For large-scale fuel ethanol manufacture, however, it is another feature of the continuous process that has proved of widest application, that is, the use of multiple fermentors linked in series with (or without) the option of recycling the fermenting broth, sometimes described as “cascading” (figure 4.9).122 Several evolutionary variants of this process paradigm have been developed by the Raphael Katzen Associates and Katzen International, Cincinnati, Ohio, from design concepts for corn ethanol dating from the late 1970s.36 In the former USSR, batteries of 6-12 fermentors were linked in multistage systems for the production of ethanol from miscellaneous raw materials in 70 industrial centers, but, unlike the West, such advanced engineering was applied to other types of fermentation, including beer, champagne, and fruit wines, as well as fodder yeast (as a form of single-cell protein); these were supported by scientists at the All-Union Research Institute of Fermentation Products who developed sophisticated mathematical analyses, unfortunately in a literature almost entirely in Russian.155

Multistage fermentations have been merged with VHG media in laboratory sys­tems with 99% of consumption of media containing up to 32% w/v glucose.156 In such a continuously flowing system, ethanol yield from glucose increased from the first to the last fermentor in the sequence at all glucose concentrations tested (15-32% w/w); ethanol was, therefore, produced more efficiently in the later stages of the inhomoge­neous set of fermentations that are set up in quasi-equilibrium within the sequence of the linked fermentation vessels. This implies mathematical modeling to control pro­ductivity will be difficult in multistage processes because parameters such as ethanol yield from sugars will be variable and possibly difficult to predict with complex feed­stocks. In similar vein, a Chinese prototype with a working volume of 3.3 l achieved a 95% conversion of glucose to ethanol; although oscillations were observed in residual glucose, ethanol concentrations, and cell densities, some success was found in devis­ing models to predict yeast cell lysis and viability loss.157 In Brazil, attention was focused on the problems of running ethanol fermentations at high ambient (tropical)

temperatures, and a five-fermentor system was devised in which a temperature gra­dient of 8°C was established: high-biomass cultures were generated at up to 43°C, and a high-viability process with continuous ethanol production was demonstrable at temperatures normally considered supraoptimal for brewer’s yeast.158 A much simpler design, one transposed from brewery work in the United Kingdom, fermented sugar­cane juice at sugar concentrations up to 200 g/l in a tower with continuous recycling of a highly flocculent yeast strain; a constant dilution rate and pH (3.3) gave outflow ethanol concentrations of up to 90 g/l with conversion rates as high as 90% of the theoretical maximum.159 In France, the concept of separate compartments for growth and ethanologenesis was pursued in a two-stage fermentor with efficient recycling of the yeast cells; a steady state could be reached where the residual glucose concentra­tion in the second stage was close to zero.160

However described — continuous, cascaded, multistage, and others — sequen­tial fermentations offer a step change in fermentation flexibility that may be of great value for lignocellulosic feedstocks: different vessels could, for example, be oper­ated at higher or lower pH, temperature, and degree of aeration to ferment different sugars and could even accommodate different ethanologenic organisms. Figure 4.10 is a schematic of a process with split pentose and hexose sugar streams with five fermentors in the continuous cascaded sequence, the liquid stream that moves from fermentor to fermentor being in contact with a stripping gas to remove the ethanol

(thus avoiding the buildup of inhibitory concentrations); the process is a composite one based on patents granted to Bio-Process Innovation, West Lafayette, Indiana.161163 A low-energy solvent absorption and extractive distillation process recovers the etha­nol from the stripping gas, and the gas is then reused, but the same arrangement of fermentation vessels can be used with conventional removal of ethanol by distillation from the outflow of the final fermentor. The pentose and hexose fermentations could, moreover, employ the same recombinant organism or separate ethanologens — an example of this was devised by researchers in China who selected Pichia stipitis as the organism to ferment pentoses in an airlift loop tower while choosing S. cerevisiae for glucose utilization in an overflow tower fermentor for the residual glucose; with an appropriate flow rate, a utilization of 92.7% was claimed from the sugars prepared from sugarcane bagasse.95