Search

A low-cost lipase preparation is required for enzymatic biodiesel synthesis. One possibility is to produce the lipase in solid-state fermentation (SSF) and then add the fermented solids (FS) directly to the reaction medium for biodiesel synthesis. In the current work, we scaled up the production of FS containing the lipases of Rhizopus microsporus. Initial experiments in flasks led to a low-cost medium containing wheat bran and sugarcane bagasse (50:50 w/w, dry basis), supplemented only with urea. We used this medium to scale-up production of FS, from 10 g in a laboratory column bioreactor to 15 kg in a pilot packed-bed bioreactor. This is the largest scale yet reported for lipase production in SSF. During scale-up, the hydrolytic activity of the FS decreased 57%: from 265 U g−1 at 18 h in the laboratory bioreactor to 113 U g−1 at 20 h in the pilot bioreactor. However, the esterification activity decreased by only 14%: from 12.1 U g−1 to 10.4 U g−1. When the FS produced in the laboratory and pilot bioreactors were dried and added directly to a solvent-free reaction medium to catalyze the esterification of oleic acid with ethanol, both gave the same ester content, 69% in 48 h.

The enzymatic route for biodiesel production is environmentally attractive, but costs need to be reduced and reaction times shortened. Costs can be reduced by producing lipases by solid-state fermentation and using the fermented solids directly to catalyze the ethyl-esterification of fatty acids in solvent-free media. However, the esterification studies done to date with fermented solids have involved 12 g or less of fermented solids and reaction times greater than 30 h. The aim of the current work was therefore to increase the scale of the process by an order of magnitude, while also decreasing the reaction time, using bi-distilled olein as the source of fatty acids. The best result was obtained with a molar ratio of ethanol to fatty acids of 1.5:1, in a process in which 245 g of ethanol and 1000 g of olein were circulated in a closed-loop system through a column containing 120 g of dry fermented solids. A two-compartment reservoir was used, with the water in the outflow of the column being decanted in the first compartment. The organic phase that passed to the second compartment was agitated and fed back to the column. This system gave 88% conversion in 24 h. This system was used in 6 successive 48-h batches with the same dry fermented solids, in which a total of 4.6 kg of ester was produced. These are promising results that lay the basis for further scale-up studies.

Solid-state fermentation could be used to produce low-cost pectinases that could then be used to saccharify pectin in citrus waste biorefineries. Recently, we produced pectinases in a pilot-scale packed-bed bioreactor, growing Aspergillus niger on a substrate mixture consisting of 27 kg of wheat bran and 3 kg of sugarcane bagasse (dry mass). However, the agglomeration of particles and shrinkage of the bed created preferential flow paths, leading to overheating within the bed and poor uniformity of pectinase levels at the end of the fermentation. In the current work, we used intermittent agitation as a strategy to minimize agglomeration, comparing one agitation (10 h), three agitations (at 8, 10 and 12 h) and five agitations (every 2 h from 8 to 16 h). The pectinase activity in the bed was uniform after agitation, but poor uniformity occurred when the bed was left unmixed for more than 10 h. The best regime was that with three agitations: For 15 samples removed from different vertical and horizontal positions of the bed at 20 h, the average pectinase activity was 22 U g−1, with a sample standard deviation of 2 U g−1. We conclude that the use of intermittently-mixed packed-bed bioreactors is a promising strategy for producing pectinases in solid-state fermentation.

Solid-state cultivation (SSC) involves the cultivation of microorganisms on moist solid particles surrounded by a continuous gas phase. Although SSC has been used for centuries in the production of traditional fermented foods, most biotechnological products are currently produced by submerged culture. However, SSC will become of increasing importance for processing solid residues in biorefineries, especially when filamentous fungi are involved. This chapter presents the basic design and operating principles of the four bioreactor types available for SSC processes: trays, packed beds, rotating (or stirred) drums, and forcefully aerated agitated bioreactors. Control of bed temperature is the main challenge, with the selection of an appropriate bioreactor depending on the specific growth rate of the process organism and also its tolerance of agitation. The bioreactors differ with respect to whether or not the substrate bed is agitated and whether or not the substrate bed is forcefully aerated, offering different combinations of conductive, convective, and evaporative cooling. The best tools for guiding the scale-up of SSC bioreactors are mathematical models that integrate growth kinetics with energy and water balances and which recognize and describe the gradients that occur across the bed during periods of static operation. Due to the complexity of the microscopic-scale phenomena involved in growth of the microorganism on the substrate particles, the kinetic equations used in these models are usually simple empirical equations, such as the logistic equation.

In this review, we describe our experience in building a pilot-scale packed-bed solid-state fermentation (SSF) bioreactor, with provision for intermittent mixing, and the use of this bioreactor to produce pectinases and lipases by filamentous fungi. We show that, at pilot scale, special attention must be given to several aspects that are not usually problematic when one works with laboratory-scale SSF bioreactors. For example, it can be a challenge to produce large amounts of inoculum if the fungus does not sporulate well. Likewise, at larger scales, the air preparation system needs as much attention as the bioreactor itself. Sampling can also be problematic if one wishes to avoid disrupting the bed structure. In the fermentations carried out in the pilot bioreactor, when the substrate bed contained predominantly wheat bran, the bed shrank away from the walls, providing preferential flow paths for the air and necessitating agitation of the bed. These problems were avoided by using beds with approximately 50% of sugarcane bagasse. We also show how a mathematical model that describes heat and water transfer in the bed can be a useful tool in developing appropriate control schemes. Graphical Abstract.

We review the application of dry fermented solids (DFS) containing naturally immobilized enzymes as catalysts in synthesis and in hydrolysis reactions. The most studied application is the use of DFS containing lipases in the synthesis of biodiesel esters, by transesterification of oils or by esterification of fatty acids with short-chain alcohols in solvent-free reaction media. Other applications of DFS that have been studied include the use of DFS containing pectinases to liberate D-galacturonic acid from pectin and the production of high-value compounds by DFS containing lipases, such as the synthesis of sugar esters and the production of pure enantiomers by resolution of racemic mixtures. To date, studies are limited to proof of concept, and there are still many challenges to be faced in the development of industrial-scale processes using DFS as catalysts. A key challenge is the relatively low activity of DFS compared to commercial enzyme preparations. Attention needs to be given to scale up, not only of the bioreactor for the application of the DFS but also for the production of the fermented solids. There is also a need for economic feasibility studies to determine whether the production of DFS and their use as catalysts can be competitive at industrial scale. Graphical Abstract.

The high cost of lipases is a barrier to producing biodiesel through the enzymatic route. Costs can be reduced by producing the lipases by solid-state fermentation and simply drying the fermented solids to produce the biocatalyst. When dry fermented solids (DFS) are used to catalyze the ethyl esterification of fatty acids, they preferentially sorb hydrophilic components of the reaction medium, thereby complicating description of the reaction kinetics. Here, we use a classical equation from heterogeneous catalysis, the two-site Langmuir-Hinshelwood-Hougen-Watson (LHHW) equation, to describe the ethyl esterification of oleic acid catalyzed by DFS, using Aspen Plus®. The parameters were determined using literature data for the ethyl esterification of bidistilled olein in a closed-loop system with recirculation of reaction medium from a well-mixed reservoir through a packed-bed bioreactor containing DFS. The model was validated with data obtained for ethyl esterification in the same bioreactor, but with the following changes: (1) use of fatty acids from soybean soapstock acid oil as the feedstock and (2) use of bidistilled olein as the feedstock, but with a reservoir with a separation barrier, so that only the organic phase is recirculated through the bed. We conclude that the two-site LHHW equation is suitable for modeling the ethyl esterification of fatty acids in techno-economic analyses of biodiesel production catalyzed by DFS that contain lipases.

Although mathematical models of heat and water transfer in packed-bed bioreactors for solid-state fermentation have been available for several decades, there have been few efforts to determine the heat and mass transfer coefficients between the solids and air phases experimentally. In the current work, we undertook drying, cooling and heating experiments in a pilot packed-bed bioreactor, using a solid substrate prepared as for a fermentation, but uninoculated. We then used the solids-air heat and mass transfer coefficients as fitting parameters to adjust a heat and mass transfer model to temperature data obtained at different bed heights. The model fit the data well, but only if the air flow rate was allowed to vary as a function of position, indicating that the air flow in the bed was not uniform. In a sensitivity analysis, the air flow rate had a greater effect on the predicted temperature profiles than did the solids-air heat and mass transfer coefficients themselves. However, despite the lesser effect of these coefficients, it is not possible to treat the solids and air phases as being in thermal and hydric equilibrium. Rather, it is essential to use models that recognize the solids and air as separate phases and use expressions based on driving forces to describe evaporation and convective heat removal. The experimentally determined heat and mass transfer coefficients can be substituted into a mathematical model that also describes microbial growth and metabolic heat production and this model can be used to guide the scale-up from our pilot bioreactor to larger scales.

Model-based strategies for scaling up traditional packed-bed bioreactors used in solid-state fermentation have been available for over two decades. However, these strategies are theoretical, not being based on experimental data. A recently proposed strategy did use experimental data; however, it was based on data from a slow-growing fungus and also considered heat transfer in both the axial and radial directions, ignoring the fact that radial heat transfer is negligible in large-scale packed-bed bioreactors. The current work demonstrates a scale-up strategy based on a model describing heat transfer only in the axial direction. The strategy is demonstrated for the cultivation of Aspergillus niger, a fast-growing filamentous fungus. In the model, the kinetics of production of metabolic heat are based on an experimental time-course profile for the O2 uptake rate. The model is validated by comparing its predictions for temperature profiles in the bed with temperatures measured experimentally in a pilot-scale bioreactor. Finally, the validated model is used to explore strategies for optimizing the performance of large-scale traditional packed-bed bioreactors. We conclude that a key strategy is to decrease the temperature of the inlet air during periods of peak heat generation.

Pectinases can be used in citrus waste biorefineries to hydrolyze the pectin in citrus pulp to produce d -galacturonic acid, a potential platform chemical. Solid-state fermentation has the potential to produce low-cost pectinases for such biorefineries, but it is difficult to control the process at large scales. In the current work, Aspergillus oryzae was cultivated in a pilot-scale packed-bed bioreactor, on 15 kg of a substrate containing 51.6% citrus pulp and 48.4% sugarcane bagasse (w/w, dry basis). The sugarcane bagasse gave a high bed porosity and ensured a stable bed structure, avoiding problems of bed shrinkage and the formation of compact agglomerates within the bed. As a result, bed temperatures were controlled to within 1 °C of the inlet air temperature and pectinase yields of 33–41 U g −1 were obtained across the bed. When the fermented solids were dried and added directly to a pectin solution, they gave a profile for the release of d -galacturonic acid similar to that obtained with a commercial pectinase. These results show the potential for using solid-state fermentation to produce pectinases in a citrus waste biorefinery, with subsequent direct addition of the fermented solids to produce d -galacturonic acid from the pectin contained in the citrus pulp.

Solid-state fermentation can be used to produce pectinases using agro-industrial byproducts. However, heat and mass transfer limitations make it difficult to control the temperature within the bioreactor, especially at large scale, so reliable scale-up strategies are essential. In the current work, we scaled up the production of pectinases in packed-bed bioreactors, from 12 g to 30 kg of dry substrate, the biggest scale yet reported for pectinase production. When compaction occurred, bed temperatures up to 47 °C were recorded and the pectinase activity in different regions of the bed at 26 h varied from 11 to 28 × 10 3 U kg −1 . When compaction was avoided, the maximum bed temperature was 32 °C and the pectinase activity at 26 h varied from 17 to 20 × 10 3 U kg −1 . The best result was obtained with a 40-cm high bed containing 27 kg of wheat bran and 3 kg of sugarcane bagasse, with switching of the temperature of the saturated inlet air between 24 °C and 32 °C. Under these conditions, the maximum productivity was 1840 U kg −1 h −1 at 10 h. We propose that the process can be scaled up to production scale by maintaining the same bed height and operational strategy, while increasing the width of the bed to several meters. If the superficial velocity of the air is maintained constant at 0.1 m s −1 , then the performance at scales involving several tonnes of solid substrate should be similar to that obtained in pilot-scale bioreactor in the current study.

A fermented solid containing lipases was produced by growing Burkholderia cepacia LTEB11 on a 1:1 mixture, by mass, of sugarcane bagasse and sunflower seed meal. This fermented solid was dried and then used in a packed-bed bioreactor to catalyze the esterification of fatty acids from soybean soapstock acid oil with ethanol in a solvent-free system. The bioreactor was operated in batch mode, with the medium being recirculated from the reservoir through the bed in a closed-loop system. During the reaction, the bulk reaction medium separated into organic and aqueous phases. Additionally, up to 30% of the reaction medium was held as a sorbed phase on the fermented solids. The composition of this sorbed phase, which represents the microenvironment of the lipases, was significantly different from the compositions of the bulk organic and aqueous phases. It was predominantly polar, with water or ethanol as the major component, but also contained significant amounts of ester. Importantly, the molar ratios of ethanol to fatty acid in the sorbed phase were significantly higher than those in the bulk reaction medium; this has implications for the reaction, since ethanol is known to cause inhibition and denaturation of lipases.

We investigated a new hydroesterification strategy for the production of biodiesel from low-value oil feedstocks: complete hydrolysis of the feedstock to fatty acids in subcritical water, followed by the use of a packed-bed reactor, containing a fermented solid with lipase activity, to convert the fatty acids to their ethyl esters. The fermented solids were produced by cultivating Burkholderia cepacia LTEB11 for 72 h on a 1:1 mixture, by mass, of sugarcane bagasse and sunflower seed meal. The esterification of fatty acids obtained from soybean soapstock acid oil was studied in the packed-bed bioreactor, in a solvent-free system, with the best results being a 92% conversion in 31 h, obtained at 50 °C. When the packed-bed reactor was reused in successive 48-h esterification reactions, conversions of over 84% of the fatty acids to esters were maintained for five cycles at 50 °C and for six cycles at 45 °C. Unlike previous hydroesterification processes that have used lipase-catalyzed hydrolysis followed by chemically-catalyzed esterification, our process does not expose the lipases to contaminants present in low quality feedstocks such as soapstocks. This advantage opens the possibility of operating the packed-bed esterification reactor in continuous mode.

We use a mathematical model, based on the N-tanks-in-series approach, to evaluate the potential advantages that can be obtained in solid-state fermentation processes by operating packed-bed bioreactors as multi-layer beds. We explore classical operation, in which air is blown unidirectionally through a substrate bed that remains immobile, with two strategies that involve movement of the layers. The first strategy involves batch operation, in which the positions of the layers are changed at 1 h intervals, in a cycling motion. The second strategy involves continuous plug-flow of the layers, with the regular addition of new layers at the air outlet and removal of spent layers at the air inlet. Under the conditions of the simulation, the rate of metabolic heat generation during the steady state of the continuous plug-flow process is only 60% of the peak value predicted for classical operation. As a result, the maximum bed temperature in the continuous plug-flow process is 4.5 °C lower than that predicted for classical operation. We conclude that the operation of multi-layer packed beds in the continuous plug-flow mode can improve bioreactor performance significantly.

Pseudomonas aeruginosa UFPEDA 614 produced a rhamnolipid biosurfactant when grown on sugarcane bagasse impregnated with a solution containing glycerol. Biosurfactant levels reached 40 g of rhamnolipid per kilogram of dry initial substrate after 12 days. On the basis of the volume of liquid used, the biosurfactant levels were similar to those obtained in submerged liquid culture of a medium identical to the impregnating solution. The properties of the biosurfactant were very similar to those obtained with rhamnolipids produced in submerged culture, with a critical micelle concentration of 46.8 mg/L and an emulsification index at 24 h of over 90% against gasoline. The surface properties were maintained after autoclaving of the fermented solids, meaning that it is possible to minimize safety risks by killing the producing organism with a heat treatment of the solids prior to product extraction. The biosurfactant was used in the washing of soils contaminated with gasoline. An aqueous biosurfactant solution was 3.2-fold more efficient than water in leaching organic material from the soil, demonstrating the viability of application of rhamnolipids in the bioremediation of soils contaminated with gasoline.

Solid-state fermentation (SSF) with filamentous fungi is a promising technique for the production of a range of biotechnological products and has the potential to play an important role in future biorefineries. The performance of such processes is intimately linked with the mycelial mode of growth of these fungi: Not only is the production of extracellular enzymes related to morphological characteristics, but also the mycelium can affect bed properties and, consequently, the efficiency of heat and mass transfer within the bed. A mathematical model that describes the development of the fungal mycelium in SSF systems at the particle scale would be a useful tool for investigating these phenomena, but, as yet, a sufficiently complete model has not been proposed. This review presents the biological and mass transfer phenomena that should be included in such a model and then evaluates how these phenomena have been modeled previously in the SSF and related literature. We conclude that a discrete lattice-based model that uses differential equations to describe the mass balances of the components within the system would be most appropriate and that mathematical expressions for describing the individual phenomena are available in the literature. It remains for these phenomena to be integrated into a complete model describing the development of fungal mycelia in SSF systems.

We present a model for the kinetics of microbial growth that incorporates the influence of the temperature variations that typically occur during solid-state culture in large-scale bioreactors. The model proposes that the specific growth rate constant depends on the level of an essential component within the biomass, with the rate of the synthesis and denaturation reactions of this component depending on temperature according to the Arrhenius equation. Model predictions were compared to literature data for the growth of Rhizopus oligosporus in three different situations: (1) incubation of cultures at different but constant temperatures; (2) initial incubation of cultures at 37 °C, followed by incubation at 50 °C; and (3) tray culture. The model agreed reasonably with all three sets of experimental results with the use a single set of parameter values. This approach to modeling has good potential for application in models of solid-state culture bioreactors.

The potential for thermal denaturation to cause enzyme losses during solid-state fermentation processes for the production of enzymes was examined, using the protease of Penicillium fellutanum as a model system. The frequency factor and activation energies for the first-order denaturation of this enzyme were determined as 3.447 × 10 59 h −1 and 364,070 J mol −1 , respectively. These values were incorporated into a mathematical model of enzyme deactivation, which was used to investigate the consequences of subjecting this protease to temporal temperature profiles reported in the literature for mid-height in a 34 cm high packed-bed bioreactor of 150 mm diameter. In this literature source, temperature profiles were measured for 5, 15 and 25 liters per minute of air and enzyme activities were measured as a function of time. The enzyme activity profiles predicted by the model were distributed similarly, one relative to the other, as had been found in the experimental study, with substantial amounts of denaturation being predicted when the substrate temperature exceeded 40 °C, which occurred at the lower two airflow rates. A mathematical model of a well-mixed bioreactor was used to explore the difficulties that would be faced at large scale. It suggests that even with airflows as high as one volume per volume per minute, up to 85% of the enzyme produced by the microorganism can be denatured by the end of the fermentation. This work highlights the extra care that must be taken in scaling up solid-state fermentation processes for the production of thermolabile products.

Mathematical models are important tools for optimizing the design and operation of solid-state fermentation (SSF) bioreactors. Such models must describe the kinetics of microbial growth, how this is affected by the environmental conditions and how this growth affects the environmental conditions. This is done at two levels of sophistication. In many bioreactor models the kinetics are described by simple empirical equations. However, other models that address the interaction of growth with intraparticle diffusion of enzymes, hydrolysis products and O2 with the use of mechanistic equations have also been proposed, and give insights into how these microscale processes can potentially limit the overall performance of a bioreactor. The current article reviews the advances that have been made in both the empirical- and mechanistic-type kinetic models and discusses the insights that have been achieved through the modeling work and the improvements to models that will be necessary in the future.

Mathematical models are important tools for optimizing the design and operation of solid-state fermentation (SSF) bioreactors. Such models must describe the transport phenomena within the substrate bed and mass and energy exchanges between the bed and the other subsystems of the bioreactor, such as the bioreactor wall and headspace gases. The sophistication with which this has been done for SSF has improved markedly over the last decade or so. The current article reviews these advances, showing how the various transport phenomena have been modeled. It also discusses the insights that have been achieved through the modeling work and the improvements to models that will be necessary in order to make them even more powerful tools in the optimization of bioreactor performance.

Over the last decade there has been a significant improvement in understanding how to design, operate and scale-up solid-state fermentation bioreactors. The key to these advances has been the application of mathematical modeling techniques to describe the biological and transport phenomena within the system. This review focuses on the advances in understanding that have come from this modeling work, and the insights it has given us into bioreactor design, operation and scale-up. It also highlights two promising bioreactor designs that have emerged over the last decade or so. For processes in which the substrate bed must remain static throughout the fermentation, the most promising design is the Zymotis design of ORSTOM at Montpellier, France, which involves closely spaced internal heat transfer plates within a packed-bed bioreactor. For those processes in which mixing can be tolerated, the stirred bioreactor developed at INRA, in Dijon, France, has been successfully demonstrated at scales of 1–25 t of substrate. Theoretical work suggests that mathematical models will be useful tools in the scale-up process, however, there are no reports that they have been used in the development of any current large-scale process. Rather, the models have been validated against data obtained from laboratory-scale bioreactors. There is an urgent need to test the accuracy and robustness of the models by applying them within real process development.

In recent years biosurfactants have attracted attention because of their low toxicity, biodegradability and ecological acceptability. However, their use is currently extremely limited due to their high cost in relation to that of chemical surfactants. Solid-state cultivation represents an alternative technology for biosurfactant production that can bring two important advantages: firstly, it allows the use of inexpensive substrates and, secondly, it avoids the problem of foaming that complicates submerged cultivation processes for biosurfactant production. In this chapter we show that, despite its potential, to date relatively little attention has been given to solid-state cultivation for biosurfactant production. We also note that this cultivation technique brings its own challenges, such as the selection of a bioreactor type that will allow adequate heat removal, of substrates with appropriate physico-chemical properties and of methods for monitoring of the cultivation process and recovering the biosurfactants from the fermented solid. With suitable efforts in research, solid-state cultivation can be used for large-scale production of biosurfactants.

Solid-state fermentation is a cultivation technique in which microorganisms are cultivated on moist solid particles, in beds within which there is a continuous gas phase between the particles. This cultivation technique will become of increasing importance in the future, as population growth puts an ever greater pressure on world resources, stimulating the development of biorefineries: the use of solid-state fermentation in many of the biological processing steps in the biorefinery will help to minimize water use. However, in order to fulfill this potential, it will be necessary to have reliable large-scale solid-state fermentation bioreactors and strategies for optimizing their operation. This article evaluates the current state of our knowledge in this area. It presents the various bioreactors that have been used to date, classifying them on the basis of aeration and agitation strategies: trays, packed beds, rotating/stirred drums, and forcefully aerated agitated bioreactors. It gives an overview of the state of the art in the application of engineering principles to the design and operation of these bioreactor types, identifying where further work is necessary.

In recent years biosurfactants have attracted attention because of their low toxicity, biodegradability and ecological acceptability. However, their use is currently extremely limited due to their high cost in relation to that of chemical surfactants. Solid-state cultivation represents an alternative technology for biosurfactant production that can bring two important advantages: firstly, it allows the use of inexpensive substrates and, secondly, it avoids the problem of foaming that complicates submerged cultivation processes for biosurfactant production. In this chapter we show that, despite its potential, to date relatively little attention has been given to solid-state cultivation for biosurfactant production. We also note that this cultivation technique brings its own challenges, such as the selection of a bioreactor type that will allow adequate heat removal, of substrates with appropriate physico-chemical properties and of methods for monitoring of the cultivation process and recovering the biosurfactants from the fermented solid. With suitable efforts in research, solid-state cultivation can be used for large-scale production of biosurfactants.

Solid-state cultivation involves the growth of microorganisms in beds of moist solid particles that have a minimum of free water between the particles. The chapter describes environmentally-related solid-state cultivation processes. For example, some processes use substrates that are residues of agriculture, forestry, or food-processing, thereby reducing the environmental impact of the residue. Other processes do not use residues, but produce products that have environmental applications. Still other processes use environmental-friendly biotransformations that have the potential to replace current industrial processes. Finally, some solid-state cultivation processes can be used to remove pollutants from soil or waste streams. Typically, environmental applications of solid-state cultivation involve large-scale processing of organic solids. The current chapter addresses the design and operation of bioreactors for these processes. It shows how the various bioreactor types can be classified according to the aeration strategy, namely whether the bed of solids is forcefully aerated or not, and according to the agitation strategy, namely the frequency of mixing of the bed of solids. It discusses the current state-of-the-art in optimizing the design and operation of the various bioreactor types, showing how mathematical models that combine microbial growth kinetics and heat and mass transfer phenomena are the most powerful tools that we have available for this task. The chapter concludes by highlighting the necessity to convert current mathematical models into user-friendly computer programs that can guide design and operation decisions for large-scale solid-state cultivation bioreactors.

This chapter addresses the design and operation of rotating-drum bioreactors and those stirred-drum bioreactors in which the air is blown into the headspace and not forcefully through the substrate bed itself. This type of bioreactor might be chosen for continuous processes, which will be discussed in Chap. 11. It can also be used for batch processes, which will be the focus of this chapter. Note that there are several bioreactors that are very similar in appearance to rotating-and stirreddrum bioreactors in which air is introduced directly into the bed. This forced aeration would tend to place them in the Group IVa bioreactors considered in Chap. 9 (continuously-agitated, forcefully-aerated bioreactors), however, whether the bioreactor performs more closely to this type of bioreactor or to a rotating-drum or stirred-drum bioreactor depends on the effectiveness of this forced aeration.

The layer of substrate in trays is limited to a bed height of around 5 cm by considerations of heat and O2 transfer within the bed. Therefore scale-up of the process cannot be achieved by increasing the bed height. The only manner to scale up a tray process to large scale is to increase the surface area of the trays, which is equivalent to saying that the large-scale process must use a large number of trays of the same size as those in which the laboratory studies were done. The use of large numbers of trays implies the necessity either for manual handling or highly sophisticated robotic systems, both of which can be inordinately expensive. However, in regions in which manual labor costs are low, such tray-type processes may find applications.

True packed-bed operation can only be used in those cases where the bed does not dry out to levels that cause water limitations on growth, because water can only be uniformly distributed within a bed of solids while the solids are being agitated. If the organism tolerates some mixing, then the intermittently-mixed mode of operation can be used, in which the bioreactor operates as a packed-bed during the majority of the fermentation period and undergoes infrequent mixing events, during which water can be added to the bed (Chap. 10). In fact, once the intermittentlymixed mode of operation is selected, the use of dry air to promote evaporative cooling is potentially available as an operating strategy. The current chapter presents a model that can be used to investigate the operation of such bioreactors.

So far we have addressed the graphical and mathematical issues associated with constructing and analyzing the kinetic profile. The next chapter gives advice about the experimental techniques that may need to be used.

As argued in Chap. 1, mathematical models of bioreactor operation will be important tools in the development of bioreactors for solid-state fermentation (SSF) processes. These mathematical models must describe quantitatively the various phenomena within the SSF process that can potentially limit the performance of the bioreactor. One of the key early steps in modeling is to identify what these phenomena are, and to unite them in a qualitative description of the system, at an appropriate level of detail (an idea that will be developed further in Chap. 13). The current chapter provides a basis for this by describing SSF processes qualitatively, from several different perspectives. The current chapter presents

While attention should certainly be given to furthering the development of fully-mechanistic models, at the moment fast-solving models are sufficiently accurate to be useful tools in the design of bioreactors and the optimization of their operation. The rest of the book concentrates on fast-solving models. Chapters 14 to 17 describe approaches to establishing and modeling the growth kinetics in a manner appropriate for incorporation into fast-solving models. Chapters 18 to 20 show how the heat and mass transfer phenomena within bioreactors can be described at an appropriate level of detail for a fast-solving model. Chapters 22 to 25 then present several fast-solving models and show how they can be used to give insights into optimal design and operation. We are confident that readers, with relatively little effort, can adapt these models to their own systems, and obtain useful results from doing so.

Intermittently-mixed, forcefully-aerated bioreactors appear to have some potential, judging by the fact that several processes involving bioreactors that operate in this mode have been demonstrated at a reasonably large scale. They appear to offer some benefits in control of the conditions within the bed, while minimizing the deleterious effects that continuous mixing can have, at least for fungal processes.

Chapters 14 to 17 have given an overview of the experimental and mathematical steps in the development of the kinetic sub-model of an SSF bioreactor model. Chapters 18 to 20 will address basic principles related to the balance/transport sub-model.

This chapter presents a case study to show how modeling work can provide insights into how to best design and operate well-mixed rotating and stirred drum bioreactors. Recently, Schutyser et al. (2001, 2002, 2003c) have developed more sophisticated models for rotating-drum bioreactors. These models describe the movement of individual particles during drum operation. Although they are potentially quite powerful tools for exploring bioreactor behavior, they are much more complex to set up, and solution times can be as long as several days.

Despite centuries of use and renewed interest over the last 20 years in solid-state fermentation (SSF) technology, and despite its good potential for a range of products, there are currently relatively few large-scale commercial applications. This situation can be attributed to the complexity of the system: Macroscale and microscale heat and mass transfer limitations are intrinsic to the system, and it is only over the last decade or so that we have begun to understand them. This review presents the current state of understanding of biochemical engineering aspects of SSF processing, including not only the fermentation itself, but also the auxiliary steps of substrate and inoculum preparation and downstream processing and waste disposal. The fermentation step has received most research attention. Significant advances have been made over the last decade in understanding how the performance of SSF bioreactors can be controlled either by the intraparticle processes of enzyme and oxygen diffusion or by the macroscale heat transfer processes of conduction, convection, and evaporation. Mathematical modeling has played an important role in suggesting how SSF bioreactors should be designed and operated. However, these models have been developed on the basis of laboratory-scale data and there is an urgent need to test these models with data obtained in large-scale bioreactors.

We studied the production of surfactin by Bacillus pumilus UFPEDA 448 in solid-state fermentation (SSF), using a medium based on okara with the addition of sugarcane bagasse as a bulking agent. The optimum proportions of okara and sugarcane bagasse were 50% each, by mass. Due to the relatively high production of proteases during SSF, pre-hydrolysis of the okara with a protease did not improve surfactin levels. The optimum temperature for surfactin production was 37 °C, with the incubation temperature affecting the ratios of the various surfactin homologues produced. Cultivation in column bioreactors with forced aeration under optimized conditions gave a surfactin level of 809 mg L −1 of impregnating solution, which corresponds to 3.3 g kg-dry-solids −1 . This is the highest surfactin level that has been produced to date in SSF with a non-recombinant microorganism. The peak O 2 uptake rate was 20 mmol min −1 kg-initial-dry-solids −1 , corresponding to metabolic waste heat production rate of 182 W kg-initial-dry-solids −1 . The tensioactive properties of the surfactin were similar to those reported in the literature for surfactin produced by submerged fermentation. These results suggest that it might be feasible to use SSF to produce surfactin but at large scale special attention will need to be given to heat removal.

Two approaches are compared for scale-up of solid-state fermentation processes in packed-bed bioreactors, one based on a dynamic heat transfer model, and the other based on a modified Damkohler number. A critical bed height is proposed, being the maximum bed height which can be used without undesirable temperatures being reached in the substrate bed during the fermentation. It depends on the microbial specific growth rate, as well as the superficial velocity and inlet temperature of the air. The critical heights predicted by the two approaches are almost identical, suggesting that approaches to scaling-up packed-bed bioreactors based on the modified Damkohler number may be successful. The modified Damkohler number is then used to predict how simple rules of scale-up might perform. Superficial velocities will need to increase with scale, although for practical reasons, it is not possible to increase superficial velocity in direct proportion to height indefinitely. A strategy is proposed to guide experimental programs for scaling-up of packed-bed bioreactors.