Industrial fermentation processes

Many Products are made by large scale fermentation today, including amino acids, enzymes, organic acids, vitamins, antibiotics, solvents and fuels (Table 53.1). There are several advantages to making products by fermentation: _ Complex molecules that occur naturally, such as antibiotics, enzymes, and vitamins, are impossible to produce chemically.   _ Optically active compounds, such as amino acids and organic acids, are difficult and costly to prepare chemically. _ “Natural” products that can be economically derived by chemical processes, but, for food purposes, are better produced by fermentation, such as beverage ethanol and vinegar (acetic acid). _ Fermentation usually uses renewable feedstocks instead of petrochemicals. _ Reaction conditions are mild, in aqueous media, and most reaction steps occur in one vessel. _ 

The by-products of fermentation are usually environmentally benign compared to the organic chemicals and reaction by-products of chemical manufacturing. Often, the cell mass and other major by-products are highly nutritious and can be used in animal feeds. There are, of course, drawbacks to fermentation processes: _ The products are made in complex solutions at low concentrations compared to chemically derived compounds. _ It is difficult and costly to purify the product. _ Microbial processes are much slower than chemical processes, increasing the fixed costs of the process. _ Microbial processes are subject to contamination by competing microbes, requiring the sterilization of the raw materials and the containment of the process to avoid contamination. _ Most microorganisms do not tolerate wide variations in temperature and pH and are also sensitive to upsets in the oxygen and nutrient levels. Such upsets not only slow the process, but are often fatal to the microorganisms. Thus, careful control of pH, nutrients, air, and agitation requires close monitoring and control. _ Although nontoxic, the waste products are high in BOD, requiring extensive sewage treatment. Other processes considered large-scale microbial processes are industrial or municipal sewage treatment, bioremediation of contaminated soil and water, and bioleaching of metal ores. Among the largest fermentation facilities are sewage treatment plants, consisting of anaerobic primary treatment and aerobic secondary treatment facilities. Bioremediation of environmental contamination is an emerging technology. Initially, native consortia of microorganisms were used to degrade contaminants. Now, cultures are selected, developed, and grown to more rapidly rehabilitate contaminated environmental sites. 

The vast majority of large-scale fermentations use bacteria, yeast and fungi, but some processes use algae, plant, and animal cells. Several cellular activities contribute to fermentation products: _ Primary metabolites: ethanol, lactic acid, and acetic acid. _ Energy storage compounds: glycerol, polymers, and polysaccharides. _ Proteins: extracellular and intracellular enzymes, single-cell proteins and foreign proteins. _ Intermediary metabolites: amino acids, citric acid, vitamins, and malic acid. _ Secondary metabolites: antibiotics. _ Whole-cell products: single-cell protein, baker’s yeast and brewer’s yeast, bioinsecticides. Fermentation products can be growth associated or non-growth associated. Primary metabolites, such as ethanol and lactic acid, are generally growth associated, as are cell mass products, while secondary metabolites, energy storage compounds, and polymers are non-growth associated. Other products, such as proteins, depend on the cellular or metabolic function. I. HISTORY Humans have used fermentation from the beginning of recorded history to provide products for everyday use. For many centuries, most microbial processing was to preserve or alter food products for human consumption. Fermentation of grains or fruit produced bread, beer, and wine that retained much of the nutrition of the raw materials, while keeping the product from spoiling. The natural yeasts that caused the fermentation added some vitamins and other nutrients to the bread or beverage. Lactic acid bacteria fermented milk to yogurt and cheeses, extending the life of milk products. Other food products were preserved or enhanced in flavor by fermentation, such as pickled vegetables and the fermentation of tea leaves and coffee beans. Fermentation was an art until the second half of the 19th century. A batch was begun with either a “starter,” a small portion of the previous batch, or with the cultures residing in the product or vessel. The idea that microbes were responsible for fermentations was not introduced until 1857, when Louis Pasteur published a paper describing the cause of failed industrial alcohol fermentations. He also quantitatively described microbial growth and metabolism for the first time and suggested heat treatment (pasteurization) to improve the storage quality of wines. This was the first step toward sterilization of a fermentation medium to control fermentation conditions. In 1883, Emil Christian Hansen began using pure yeast cultures for beer production in Denmark. Beer and wine were produced at relatively large scale starting in the eighteenth century to satisfy the demands of growing urban populations. In the mid-nineteenth century, the introduction of denaturation freed ethanol from the heavy beverage tax burden so that it could be used as an industrial solvent and fuel. The first aseptic fermentation on a large scale was the acetone–butanol fermentation, which both Britain and Germany pursued in the years preceding World War I. 

The initial objective was to provide butanol as a precursor for butanediol, for use in synthetic rubber. After the beginning of the war, the focus of the process in Britain became acetone, which was used for munitions manufacture. Britain had previously been importing acetone from Germany. One of the people instrumental in the development of the fermentation was Chaim Weizmann, who later became the first prime minister of Israel. As the process was developed and scaled up, it was found that the producing culture, Clostridium acetobutylicum, would become overwhelmed by competing bacteria introduced from the raw materials. Thus, the culture medium had to be sterilized and the process run under aseptic conditions. All penetrations on the reaction vessel were steam sealed to prevent contamination. Production eventually took place in Canada and the United States, due to the availability of cheap raw materials. In the 1920s and 1930s, the emphasis in fermentation shifted to organic acids, primarily, lactic acid and citric acid. In the United States, where prohibition had outlawed alcoholic beverages, facilities and raw materials formerly used for alcoholic beverage production became available. 

Lactic acid is currently used as an acidulant in foods, a biodegradable solvent in the electronic industry (as ethyl lactate), and a precursor for biodegradable plastics. Citric acid is used in soft drinks, as an acidulant in foods, and as a replacement for phosphates in detergents. As part of an effort to find uses for agricultural products during the depression, the USDA Northern Regional Research Laboratory (NRRL) pioneered the use of surplus corn products, such as corn steep liquor, and the use of submerged fungal cultures in fermentation. Previously, fungi had been grown on solid media or in the surface of liquid media. This set the stage for the large-scale production of penicillin, which was discovered in 1929 in the Britain, developed in the 1930s, and commercialized in 1942 in the United States. It was the first “miracle” drug, routinely curing bacterial infections that had previously caused serious illness or death. The demand was very high during World War II and the years following. Penicillin was initially produced as a surface culture in one-quart milk bottles. The cost, availability, and handling of bottles severely limited the expansion of production. Scientists at the NRRL discovered a new production culture on a moldy cantaloupe and developed a submerged culture fermentation. This led to significant increases in productivity per unit volume and the ability to greatly increase the scale of production by using stirred tank bioreactors. The success of penicillin inspired pharmaceutical companies to launch massive efforts to discover and develop many other antibiotics in the 1940s and 1950s. Most of these fermentations were highly aerobic, requiring high aeration and agitation. As the scale of production increased, it was found that mass transfer became limiting. The field of biochemical engineering emerged as a distinct field at this time, to study mass transfer problems in fermentation and to design large-scale fermentors capable of high transfer rates. In the 1960s, amino acid fermentations were developed in Japan. Initially, L-glutamic acid, as monosodium glutamate, was produced as a flavor enhancer, to supplant MSG extracted from natural sources. Using cultures derived from glutamic acid bacteria, production of other amino acids followed. Amino acids are used in foods as nutrients, sweeteners, and flavor enhancers, and in animal feeds to increase the efficiency of low protein feeds. 

Commercial production of enzymes for use in industrial processes began on a large scale in the 1970s as well. Microbial enzymes account for 80% of all enzymes in commercial use, including grain processing, sugar production, juice and wine clarification, detergents, and high fructose corn syrup. The discovery of the tools of genetic engineering expanded the possibilities for products made by fermentation. Insulin was the first genetically engineered fermentation commercialized, developed in 1977. Since then, many genetically engineered products have been produced on a large scale. 

II. CULTURE SELECTION AND DEVELOPMENT A. Selection Microbial processes begin with the culture used for production. Once the desired product or microbial activity is defined, the selection process begins. A culture should produce, or have the metabolic potential to produce, the desired product. Other attributes are important, such as substrate specificity, growth factor requirements, growth characteristics (pH, temperature, aeration, shear sensitivity), fermentation by-products, effect of the organism on downstream processing, and environmental and health effects. Literature searches can narrow the range of cultures to be screened, saving valuable time in culture selection. Culture collections often have cultures with some or all of the desired characteristics, but it is sometimes necessary to screen cultures from the environment. Determining the type of environment in which organisms with the desired characteristics live and how to separate them from the other cultures present takes considerable care. A thorough understanding of the physiology of the desired microorganism is necessary to design a successful isolation and selection strategy. Successful isolation of a useful culture often requires a combination of several enrichment or selective methods. Also, a system of storing and cataloging potentially useful isolates is very important so that commercially viable cultures are not lost. After an initial screening, there are usually many potentially useful isolates and secondary screening is necessary to eliminate false positives and evaluate the potential of the remaining candidates. The secondary screening is usually semiquantitative or quantitative. 

The list of candidates is narrowed, as much as possible, using mass screening methods, such as agar plates with selective growth inhibitors or metabolic indicators. Large-scale screening of individual isolates in shake flask culture is very time consuming and expensive. Once a few isolates have been selected, culture and process development usually begin in parallel to condense the timeline to production. B. Development It is very rare that an isolate from the environment produces the desired product cost-effectively. Often, cultures make only minute quantities of the product. Increasing the productivity of the initial isolates requires a program of genetic improvement. Classical mutation and selection methods are used with most cultures selected from the environment because little is known about their genetics and whether the cultures possess cloning vectors such as plasmids, transposons, or temperate bacteriophage. Typically, a mutagen, such as ultraviolet light, ionizing radiation, or a chemical mutagen, is applied and the culture is grown in the presence of a selective growth inhibitor or toxin. The survivors are isolated and tested. This is usually an iterative process; mutant strains are screened, remutagenized, and reselected several times, often using higher concentrations or different selective agents, until a culture with commercial potential is obtained. Even after a process is successfully brought to production, culture improvement is ongoing to improve profitability and maintain a competitive advantage. In the past two decades, the use of genetic engineering has supplemented, and sometimes replaced, classical genetic techniques. Insertion of genes into plasmids has improved the ease with which genes from one culture are transferred to another and has allowed the production of human and other mammalian proteins, such as insulin and interferons, in microbial fermentations. Plasmids with high copy numbers and strong transcriptional promoters have dramatically improved production of many proteins and enzymes. Knowledge of enzyme structure and function has led to site-specific mutagenesis, increasing the efficiency of mutagenesis programs and reducing the deleterious effects of “shotgun” mutagenesis, used in classical genetics. The introduction of polymerase chain reaction (PCR) technology has led to the ability to genetically sample environments where isolation of cultures is extremely difficult or impossible, such as cold benthic environments barely above freezing, deep sea thermal vents where microorganisms live in temperatures well above 100 _C, or acidic hot springs, such as those in Yellowstone National Park. DNA libraries from these environments can be screened for enzymes that perform under extreme conditions without having to culture the microorganisms. 

III. PROCESS DEVELOPMENT AND SCALE-UP A. Development Process development usually overlaps culture development. The purpose of process development is the formulation of media, optimization of culture conditions, and determination of the biochemical engineering parameters used to design the full-scale bioreactors. The early stages of process development are usually performed in shake flasks, where the nutritional requirements of the culture are determined and potential media components are screened. Initial growth and production studies can be performed in shake flasks as well. The limitation of shake flask culture is that pH, oxygen content, and other environmental factors cannot be easily monitored and controlled and mass transfer studies are difficult. The next stage in process development, performed in laboratory scale fermentors, is determination of fermentation characteristics, such as pH optimum, oxygen uptake rate, growth and production rates, sensitivity to nutrients and by-products, broth viscosity, heat generation, and shear sensitivity. This information is used to determine what mode of fermentation will be used and to develop the fermentation parameters, as well as to determine the mass transfer characteristics of the fermentation used in the design of the bioreactor. In addition, the medium is developed and optimized at this stage. A variety of biochemical engineering methods for scale-up of bioreactors have been applied over the years, including constant oxygen transfer rate, constant agitation power per volume, constant impeller tip speed, equal mixing times, or similar momentum factors or feedback control to try to maintain important environmental factors as constant as possible. Each has limitations in predicting the effect of scale-up on the process.

 It is more difficult to predict how a biological process will react at the commercial scale, based on laboratory and pilot plant studies, than a chemical process. This is due to the complexity of reactions and interactions that occur in a bioreactor. There are often unforseen consequences of changing the scale of operation due to the effect of heat and mass transfer on microorganisms. For example, mixing times in a laboratory or pilot scale fermentor are a few seconds, while in large-scale fermentors, mixing times can be two minutes or more. The average conditions are the same as in a smaller fermentor, but an individual microorganism encounters a variety of suboptimal conditions for a significant period of time. Also, as the fermentor size increases, the heat generation increases proportional to the volume, while the cooling capacity increases proportional to the surface area. Therefore, larger vessels require internal cooling coils to supplement the water jacket. This can aggravate mixing problems further. The properties of the fermentation broth (viscosity, osmotic pressure, substrate, and product and waste product concentrations) and gas/liquid interactions (gas lineal velocity, surface tension, pressure gradients) are also scale dependent. Thus, fermentation scale-up is often highly empirical and based on the experiences and training of the scientists and engineers involved. The cost and risk of scale-up are higher for biological systems than for other chemical systems, due to the intermediate scale pilot plant steps required to successfully predict the outcome of full-scale operations. Some of this cost can be reduced by using existing facilities or rented facilities to test the process or by using seed fermentors (typically, 5–10% as large as production fermentors) for the final pilot plant scale. An important microbiological factor that affects the scale-up biological systems is the increase in the number of generations required for full-scale operation. Nonproducing variant strains often arise from the parent population. Most commercially used microorganisms are mutated and selected for increased product yield and rates of production. This often decreases the growth rate and hardiness of the culture. Therefore, a variant that either reverts to a previous condition or that short-circuits the selected pathways by additional mutations will have a competitive advantage. 

The percent of variants in the population in a given generation depends on the rate at which variants appear and the relative growth rate of the variant to the parent population, as expressed by the equation where Xn is the number of parent culture cells, Xm is the number of variant culture cells, _ is the ratio of the specific growth rate of the variant to the parent, _ is the rate of appearance of variants per genome per generation, and Ng is the number of generations. The effect of specific growth rate and the effect of appearance of variants on the parent strain population are shown graphically in Fig. 53.1 and Fig. 53.2, respectively. An unstable culture can cause serious disruptions to production in a large-scale fermentation plant. Variant strains are usually less sensitive to adverse conditions (extremes in temperature or pH, nutrient quality, anaerobic conditions) than parent strains, as well. A production strain that appears stable for the required number of generations under laboratory conditions may exhibit instability when introduced to a full-scale plant. It is important to test the stability of potential production strains under conditions that are as similar as possible to the conditions expected in the production environment. It is also important to establish a system of culture storage and management that minimizes the number of generations required to reach full-scale production and maintains the seed stock under stable conditions. Genetically engineered cultures have additional stability problems, which can be expressed in a similar mathematical expression as variant formation. Often, the genes for producing the desired product are located on plasmids that also include selectible markers, usually antibiotic resistance. In the laboratory, these cultures are maintained in media that contain the selective agent. It is not practical to use antibiotics in a large-scale plant; therefore, production cultures must have highly stable plasmids. The rate of plasmid loss can be much higher than variant formation in production cultures. Even if they aren’t completely lost, reduction of plasmid number can drastically reduce productivity per cell. Plasmid maintenance usually requires a compromise between the level expression of the cloned genes and culture growth rate. 

B. Media development and optimization The development of a suitable, economical medium is a balance between the nutritional requirements of the microorganism and the cost and availability of the medium components. The chemical constituents of the medium are determined by the composition of the cell mass and product, the stoichiometry of cell and product formation, and from yield coefficients, which can be estimated from shake flask or lab scale fermentor experiments. On a dry weight basis, 90–95% of microbial biomass consists of carbon, oxygen, hydrogen, nitrogen, sulfur, magnesium, and potassium. The remaining 5–10% is microelements (required in small amounts), primarily calcium, manganese, iron, copper, and zinc. The stoichiometry can be determined from the general equation: CHaOb (Carbohydrate)_A(O2)_B(NH3) →Yx/s(CH_O_N_) (Biomass) _Yp/s(CH_1O_1N_1)(product)_YCO2/s (CO2)_D(H2O) The terms Yx/s, Yp/s, and YCO2/s are the yield coefficients for cell mass, product and CO2, respectively. In most large-scale fermentations, the carbon usually comes from carbohydrates, such as sugars or starches. The nitrogen is from a variety of organic and inorganic nitrogen sources and from ammonia added to control the pH. Phosphate, sulfur, and magnesium are added as salts or in complex nutrients. The micronutrients are often derived from the water or other raw materials or added as mineral salts. Some commonly used fermentation substrates are shown in Table 53.2. Many microbes are auxotrophs for specific growth factors, such as vitamins or amino acids, and will be growth limited without adequate quantities of the compound in the medium. Other cultures may not have an absolute requirement for a growth factor, but grow slowly in the absence of the specific factor. In many cases, the exact requirements are not known, but complex substrates are required for optimal growth. It is often necessary to screen a variety of potential nutrients and combinations of nutrients to satisfy the nutritional requirements and minimize the raw materials costs. Another problem encountered with some components of culture medium is seasonal or hidden auxotrophy. This is variation in growth and production encountered due to either seasonal changes in processing or with new crops of vegetable sources of raw materials. Often, this is not seen in the laboratory, but at the large scale, due to poorer mixing in large-scale fermentors. Other considerations for selection of nutrients are regulatory approval and kosher certification for some foods. 

IV. LARGE-SCALE OPERATION A. Inoculum production The starting place for large-scale fermentations is in the inoculum laboratory. The fermentation is doomed to failure without pure, active inoculum. This starts with the storage of culture. It is important to store the culture under conditions that retain both genetic stability and viability. A variety of methods have been used, all with advantages and disadvantages. Storing cultures in agar stabs or on agar slants is one of the oldest methods. It is simple and cultures can be stored at room temperature or under refrigeration, but viability and stability are limited and cultures must be transferred frequently. Lyophilization (freeze drying) requires no refrigeration or freezing and cultures can be stored almost indefinitely at room temperature and can serve as a long-term method of safely storing culture in the event of catastrophe. Lyophilization takes more skill and equipment than slant preparation and often has significant viability loss. Cultures are not easily revived, making it unsuitable for daily use. Cultures stored in aqueous glycerol solutions at very low temperatures, _80 _C or lower, generally have high viability and stability, are easy to produce and revive, but require cryogenic freezers or liquid nitrogen, and are, thus, susceptible to power outages and equipment failure. Once the cultures are safely stored, they must be revived, grown in the inoculum laboratory, and transferred to growth or “seed” fermentors that feed the production fermentors. Cultures are usually grown in shake flask culture in the laboratory and transferred to a suitable container for transfer to the plant. In the laboratory, culture transfers are exposed to the air, so precautions are taken to prevent incidental contamination, such as performing transfers in rooms or hoods with HEPA-filtered air and the wearing of sterile coveralls, gloves, and masks by technicians. 

Once out in the plant, all transfers are made through steamsterilized piping or hoses using differential pressure. B. Types of bioreactors There are a wide variety of of bioreactor designs (Figs. 53.3 and 53.4). Selection of a reactor design for a particular process depends on a variety of factors, including mass transfer considerations, mixing, shear sensitivity, broth viscosity, oxygen demand, reliability of operation, sterilization considerations, and the cost of construction and operation. Due to the complex nature of fermentation scale-up, a few basic reactor designs are used for most applications. The two most common are the stirred tank and the air lift reactors. Stirred tank reactors use sparged air and submerged impellers to aerate and mix the broth. They are versatile and are especially adapted to highly aerobic cultures and highly viscous fermentations. The drawbacks are high energy input and the use of rotating seals on the agitator shaft, which are a contamination risk. Even within this category, there are many variations in design, such as the style, number, and placement of impellers, the height to diameter ratio, the number and placement of coils or baffles, that affect the mixing characteristics of the vessel. Due to high risk of scale-up and the high capital cost of building large-scale fermentation facilities, most plants install stirred-tank reactors. Airlift fermentors mix the broth with air from the sparger. Some designs have an internal draft tube to direct the flow of fluid. Most airlift designs have a much greater height-to-diameter ratio than stirred tank vessels to improve oxygen transfer. The mixing is not as good as in a stirred tank but the energy input and shear forces are much lower, thus useful for shear sensitive cultures or in processes where the energy cost of agitation is a significant factor. Also at very large scale, the heat added to a stirred-tank vessel by the agitator becomes increasingly difficult to remove, adding to the cooling costs. Most large-scale airlift fermentors are used for plant effluent treatment, production of baker’s yeast, or for fungal fermentations where the size of the mycelial pellets is controlled by shear forces. In addition to the mass transfer characteristics of a fermentor, other factors are important in the operation of a bioreactor, such as the ability to clean the vessel, sterile integrity of the vessel, and maintenance costs. Some reactor designs that have excellent characteristics in the pilot plant are not practical choices for large-scale operation due to mechanical complexity that causes sterility and maintenance problems on scale-up. C. Modes of operation There are three basic modes of operation: batch, fed batch, and continuous, with variations of these three basic modes (Fig. 53.5). In batch mode, all the ingredients required for fermentation except pH control chemicals, usually ammonia, are added to the fermentor prior to inoculation. 

The fermentation is run until the nutrients are exhausted, then the broth is harvested. The advantage is simplicity of operation and reduced risk of contamination. It is useful in fermentations with high yield-per-unit substrate and with cultures that can tolerate high initial substrate concentrations. Fed batch mode starts with some of the nutrients in the fermentor before inoculation. Concentrated nutrients are added as the fermentation progresses. The advantages are the ability to add large quantities of nutrients to the fermentor by adding them gradually and the ability to control the rate of nutrient addition. This allows for high product concentrations without subjecting the culture to inhibition by high levels of nutrients. It also allows for control of culture growth rate, which is required in some fermentations to maximize productivity and yield. Overall, the use of the fermentor time is better in fed batch than straight batch fermentation, reducing fixed costs. The disadvantages are increased risk of contamination, due to the addition of nutrients through a continuous sterilizer, and increased equipment costs for continuous sterilization and flow control equipment for feed streams. Both batch and fed batch can be run in repeated mode, with a small portion of the previous batch left in the fermentor for inoculum. The medium is then added through a continuous sterilizer. Use of the fermentor is increased by eliminating turnaround time, but the risks of contamination and genetic degradation of the culture are increased. In any case, repeated batch mode cannot be repeated indefinitely, due to maintenance and cleaning needs. Usually, repeated fermentations are run for two or three batches. In continuous mode, the starting medium and inoculum are added to the fermentor. After the culture is grown, the fermentor is fed nutrients and broth is withdrawn at the same rate, maintaining a constant volume of broth in the fermentor. In continuous mode with cell recycle, the cell mass is returned to the fermentor using microfiltration with bacteria or screens with fungal mycelia. Continuous mode maximizes the use of the vessel and is especially good for fermentations that take a long time to reach high productivity. The disadvantages are increased risk of contamination, especially since it is difficult to keep contamination from growing through the continuous harvest line. 

As with repeated batch mode, continuous fermentations cannot be run indefinitely, but fermentations of several hundred hours’ duration are possible. D. Monitoring and control Ideally, a fermentation is best controlled using online, real-time measurements of pH, oxygen, cell mass, substrate and product concentrations, metabolic state, and nutrient flow rates and applying the data to a precise model of how the culture will respond to changes in any of the parameters. Only a few of the parameters are actually measurable in real time and the models of culture behavior are often imprecise extrapolations of experimental data, although technological advances are moving in the direction of better monitoring and control. Most physical parameters—temperature, pressure, power input, impeller speed, gas and liquid flow— can be measured accurately without invasive instrumentation. Reliable online measurement of the chemical environment has been limited to pH, dissolved oxygen, and off gas analysis. Advances in infrared analysis have recently made it possible to measure several parameters of broth composition, such as substrate, product, and ammonia concentration, simultaneously on-line. Various optical density probes, to measure cell mass, and enzyme probes, to measure nutrient or product concentration, are available, but are often subject to fouling or cannot be sterilized. Measurement of physiological characteristics, such as intracellular ATP, DNA, and RNA content, are not currently feasible and must be inferred from the physical and chemical data. At a basic level, most fermentations are run under temperature and pH control. In other fermentations, the dissolved oxygen is controlled, usually by changing the agitator speed. In a batch fermentation, not much more is usually needed. In fed batch and continuous fermentations, it is often necessary to control the nutrient concentration at a limiting concentration in order to control the growth rate of the culture, maximize product yield, or avoid anaerobic conditions. Substrate feed can be controlled using feedback control from pH, dissolved oxygen, or off gas analysis, either directly or by calculating physiological parameters, such as substrate utilization rate or oxygen uptake rate. In a pH-controlled feed system, the consumption of carbohydrate drives the pH down, and base (usually ammonia) is added to maintain a constant pH. 

When the carbohydrate becomes limiting, the pH is driven up instead. When this occurs, the feed is increased, driving the pH downward again. Sophisticated control schemes can be devised by measuring the rate of base consumption and the rate of pH increase when substrate is limiting. Without proper control loops, a pH controlled feed can bog down or overfeed, as the measurement can only tell whether there is not enough feed, not when there is too much. A dissolved oxygen (DO) controlled fermentation is maintained between two setpoints, usually between 20% and 40% dissolved oxygen, by increasing the feed when the DO rises and decreasing the feed when it drops. Without some damping of the response in DO-controlled fermentations, there is a tendency for large amplitude swings of DO and feed rate, due to delayed response and long mixing times. E. Sterilization and contamination control Sterilization is the process of eliminating all viable organisms from equipment and materials. For a fermentation plant, this requires: _ Presterilization of equipment (fermentor vessels, inoculation and feed piping, and air filters); _ Sterilization of feedstocks, either in the fermentor vessel or through a continuous sterilizer; _ Maintenance of aseptic conditions throughout the entire fermentation process. Most fermentation processes require aseptic conditions for optimal productivity and yield. Some fermentations are relatively insensitive to contamination: yeast and fungal fermentations performed at low pH, fermentations with short cycle times (high growth rates), or fermentations that produce inhibitory or toxic products (alcohols, antibiotics, organic acids) tend to be resistant to contamination, but rampant, out-of-control contamination can destroy productivity altogether. 

Other processes are highly sensitive to contamination: repeated batch or continuous fermentations and slow-growing bacterial fermentations are more susceptible to contamination and, therefore, require more stringent conditions to maintain aseptic conditions. In addition, some contaminants can be pathogenic or produce toxins that can render products unusable or be a health risk to plant workers or consumers. Therefore, complete asepsis is the operating philosophy of most largescale plants. The philosophy of aseptic operation must be an integral part of every aspect of the operation, from design and construction to operating procedures and personnel training, for a plant to operate aseptically. Most process equipment and medium components are sterilized by steam under pressure. The rate of cell death depends on the time and temperature of exposure to steam. The spores of the thermophile Bacillus stearothermophilius are the usual benchmark for determining sterilization conditions. Standard conditions for steam sterilization are 121 _C for 15 minutes, but the time required for sterilization is less at higher temperatures. Most fermentation broths and continuous feeds are sterilized by steam sterilization either in situ or in a continuous sterilization system, unless the medium contains heat-labile constituents. Sterilization of the fermentor broth in situ requires a significant amount of time in a large vessel for heating and cooling the broth, which reduces the productive time of the fermentor. Large amounts of cooling water are required as well, and the long time required for sterilization can cause degradation of some medium components. In situ sterilization, however, has the advantage of being simple, thus reducing the risk of upset in the process resulting in contamination. Continuous sterilization is sometimes used for fermentor broths but more often for nutrients added to the fermentor during the fermentation, especially for fed-batch or continuous fermentations. 

This is usually accomplished by either direct steam injection into the nutrient stream or the transfer of heat through a heat exchanger with no direct steam contact. After heating, a holding loop maintains the temperature for a specified time, followed by a second heat exchanger to reduce the heat of the medium before it is added to the fermentor. Continuous sterilization can be carried out at increased temperature and reduced time, to minimize the heat damage to the medium and reduce the length of holding tube required or increase the flow of fluid through the system. As a practical matter, heat from fluid returning from the holding loop is used, through a heat exchanger, to preheat broth going to the sterilizer, thus reducing the requirements for both steam to heat the incoming broth and cooling water to cool the sterilized broth. Duplicate systems are required for systems to be cleaned and maintained properly. Filter sterilization is used for liquids with heatsensitive components or nonaqueous liquids with low boiling points, which are added to the fermentor after sterilization and cooldown. Air and other gases are also sterilized by filtration. Two basic types of filters have been used. Depth filters consist of layers of glass wool and were mostly used to sterilize air, but can be difficult to sterilize and dry. Depth filters have been mostly replaced by absolute membrane filters, which are thin membranes with pores no larger than their rated size. Most bacteria and spores are retained by filters with 0.2_m absolute pore size. Hydrophobic filters are used to sterilize air, usually preceded by prefilters to remove solids and water from the compressed air. Bacteriophage, which are retained on dry filters by electrostatic forces, are small enough to pass through if the air is wet, making it very important to keep the process air as dry as possible. While microbial contamination usually just reduces productivity and yield, bacteriophage can destory the production culture within a few hours of introduction into the fermentor. Fermentation plant managers live in fear of bacteriophage. During the fermentation, the contents of the reactor must be isolated from potential contaminating sources. 

Besides filtering the air, all penetrations into the fermentor must be sealed from the outside. Agitators are sealed with double mechanical seals that have steam between the rotating seals. Dissolved oxygen and pH electrodes are sealed with O-rings or gaskets. Piping that penetrates the vessel, such as media fill lines, harvest lines, and inoculation lines, are isolated with steam seals, which are sections of piping between two valves that connect the fermentor to the outside “septic” environment. There is a steam source and a relief valve to vent steam condensate and maintain steam pressure above 15 psig when the line is not in use. Although valves are routinely tested for leaks, there are no valves with fine enough tolerances to prevent contamination from growing across the contact surfaces. Of all the problems encountered in a fermentation plant, contamination is the most persistent. Contamination control is a daily, ongoing activity. The culture is monitored for contamination by microscopic examination and plating on nutrient agar, from the first shake flask through the production fermentor. This helps to prevent contaminated culture from being used in production and to trace the source of contamination whenever possible. In addition to testing the culture for contamination, other potential sources of contamination, such as sterile feed streams into the fermentor, are monitored. Unfortunately, the size of sample that is practical to test for contamination is very minute in comparison to the size of the process being tested. Contamination is usually not detected until it is a serious problem. Therefore, preventive maintenance is very important to control contamination. Fermentors, piping, valves, and filters are routinely tested for integrity. Leaky valves and cracks in cooling jackets or coils are common sites for microbial contamination to enter the process, and cooling water leaks are a common source of bacteriophage contamination. Proper procedures for sterilization, culture transfer, and fermentor operation are critical to maintaining an aseptic process. This involves careful evaluation of the effect of procedures on process integrity, writing procedures clearly, training fermentor operators properly, and monitoring the effectiveness of training. 

F. Utilities Although fermentation processes usually are performed at near ambient temperatures, the energy consumption is high due to the need for sterilization, cooling, agitation, and aeration. 1. Steam Steam is required for sterilization of the medium, either in the fermentor or through continuous sterilizers. During the fermentation itself, a certain amount of steam is required to maintain the microbiological integrity of the fermentor vessel by the use of steam seals. Fermentation plants often require large quantities of steam for evaporation and drying in downstream processing, due to the low concentration of product in the broth. 2. Cooling water After sterilization, fermentors need to be cooled before they can be inoculated; during the fermentation, cultures generate great quantities of metabolic heat, and agitators add some heat to the broth. This heat must be removed to maintain the proper incubation temperature. For downstream processing, cooling water is needed for certain types of crystallizers and for condensers on evaporators. Although cooling water is recycled, there is high energy input for cooling towers and chillers and evaporative water loss in cooling towers. 3. Process water Much of the fermentor broth is water. The quality of the water used depends on the sensitivity of the culture to minerals in the water, effect of water quality on downstream processing, and regulatory requirements. In some cases, water must be purified by reverse osmosis for use in fermentation. Other fermentations can use water recycled from process condensers. Water is also used for some downstream process steps and for cleaning and rinsing the fermentors. 4. Electricity Agitation, compressed air, and water chillers require a great amount of electricity. Where economies of scale are adequate, cogeneration of steam and electricity are very cost effective. Boilers using natural gas, oil, or coal produce steam used to feed generators for electricity. The “waste” steam produced in the generation process, which is at high enough pressure for use in fermentation, is captured and used as the process steam for the plant. 5. Sewage treatment Fermentation generates waste with high BOD, mainly in the form of spent cell mass. While much of it can be used for animal feeds, some low solids streams are generated and must be treated as sewage. A lot of waste water is also generated, making the throughput for a treatment plant high. In many cases, on-site primary and secondary treatment is required to avoid paying high municipal sewage charges and fines. 

G. Downstream processing Since fermentation products are often produced in dilute aqueous solution that also includes the cell mass, metabolic by-products, and various salts, downstream processing is a major part of producing a fermentation product. The degree of purification of the product depends on the type of product and the end use. Industrial enzymes, for example, are often separated from the cell mass and concentrated by ultrafiltration to make a crude extract. On the other extreme, antibiotics for human use undergo many steps of purification. Common methods include microfiltration and ultrafiltration to remove cell mass and other debris or retain larger molecular weight proteins; concentration by evaporation or reverse osmosis; and crystallization by chilling or evaporation, pH or solvent precipitation, centrifugation, and chromatography.