化学合成生物（英文版）：05618_04

58 Single cell protein 4.1 Introduction 60 4.2 Conventional protein sources 60 4.3 Single cell protein 62 66 67 69 4.4 Substrates for SCP production 4.5 Micro-organisms for SCP production 4.6 SCP from carbon dioxide 4.7 SCP from carbohydrates 74 4.8 SCP from hydrocarbons and derivatives 85 88 4.9 The Pruteen process - a case study Summary and objectives 106 Resource Material 107 60 Chapter 4 Single cell protein 4.1 Introduction In this chapter we examine the processes that have been developed to produce micro-organisms as a source of food protein. We will examine the reasons why micmrganisms have been considered as alternative protein sources, the substrates on which they have been grown, the various process technologies developed and the comparative economics of these processes. One process will be examined in depth, to illustrate how a team composed of such diverse people as microbiologists, process engineers, patent lawyers and cost analysts work together to develop a marketable product. The driving forces behind the development of many single cell protein projects emerged from global economic conditions and social concerns of the 1960s. In the 1970s and early 198Os, there were considerable technological advancements associated with single cell protein process developments and many types of processes were operated commercially. In this chapter we present technological and economic data derived from these early developments to provide a historical context for single cell protein as a food and animal feed source. You will see that many important principles underpinning modem process technology are based on the experiences gained in the development of single cell protein processes. 4.2 Conventional protein sources essential aminoacids Animals, including humans, cannot synthesise all the different amino acids they need and thus require them in their diet. These amino acids are called the essential amino acids. Proteins in food are hydrolysed in the digestive tract and the resulting amino acids are reassembled into proteins within the animal's cells. All animals are ultimately dependent on plants for protein, as it is plants that create protein by combining inorganic nitmen from the soil (as nitrate) with organic molecules derived from carbon from the atmosphere (as COJ. For us to remain perfectly healthy, the protein in our diet must supply suffiaent quantities of amino acids. We prefer to eat our protein in particular forms, that is in foods having particular textures, tastes and smells (these are called organoleptic properties). Conventional sources of protein are plants, mainly as cereals and pulses, and animals, mainly as meat, eggs and milk. The proportions of such proteins eaten in various parts of the world differ widely (Figure 4.1). organoleqtic PmeS Single cell protein 61 Figure 4.1 World protein consumption List three factors you think account for such variations in the sources of proteins n between various parts of the world? Essentially the answer is history, climate, culture and money! Historically, people had to eat the food available locally, and this would be mntrolled by the local natural 62 Chapter 4 changing demands for dietary protein single cell and fwd proteinlfood filamentous blue-green bederia environment. Cultural influences have also led to preferences for certain food types. In more affluent countries foods such as meat, or high-protein feedstuffs on which to rear animals, can be produced or be imported. In less affluent countries such luxuries cannot be afforded, Increasing populations in some countries have overstretched food supplies, and so limited the availability of foods. There are problems, however, with these conventional sources of protein. Crop production is dependent upon a suitable climate and in most countries available arable land is already fully farmed. Fish stocks in the oceans are in danger of becoming depleted. In countries where animal meat forms a high proportion of dietary protein, there are controversies such as whether or not the fats eaten with the protein are healthy, whether or not we are justified in keeping animals in the unnatural conditions of some farms, or whether or not we are justified in killing animals for food at all. Such controversies are leading an increasing number of people to become vegetarian. It is likely that the world’s population will double in the next few decades, yet the United Nations estimate that about one thousand million people are already suffering protein deficiency. It is estimated that between 1980 and 2000 the annual demand for protein as food for humans will increase from 50 x lo6 tonnes to 79 x lo6 tonnes, and the demand for protein as feed for animals will increase from 44 x lo6 tonnes to 108 x lo6 tonnes. Biotechnology is being applied to the rapid improvement of conventional food sources, both plant and animal, in an effort to meet the increased demand in food. Interest has also been shown in growing micro-organisms as a source of protein and it is developments in this area that we are going to examine here in detail. 4.3 Single cell protein Single cell protein, normally called simply s8, is the term used to describe microbial cells, or proteins from them, which are used as food (food for humans) or feed (food for farm animals or fish). Although the term micro-organisms covers viruses, bacteria, fungi, algae and protozoa, viruses and protozoa are not considered suitable for SCP production. n Why do you think viruses and protozoa are not suitable for SCP production? Both viruses and protozoa are difficult to grow in culture. Viruses need living cells to grow in and their small size makes them difficult to deal with. Protozoa need complex diets of organic materials. Bacteria, fungi and algae are relatively easy to grow in culture. The term SCP is not exactly appropriate, as some filamentous organisms are used as SCP and these organisms are multicellular not unicellular. You may be wondering why anyone should ever have considered using micmaganisms as a protein source. Let us consider why this should have been. 4.3.1 The advantage of micro-organisms as a protein source Eating micro-organisms is nothing new. You might not have been aware that some foods traditionally eaten by man are in fact micro-organisms. Filamentous blue-p;reen bacteria (often called blue-green algae, or cyanobacteria) were collected from lakes and rivers and eaten by the Aztecs in Mexico, and people inhabiting the shores of Lake Chad in Africa still do so. Edible fungi have been collected from the wild for centuries and farmed throughout the last two. During the two World Wars this century, yeasts (unicellular fungi) were grown on a large scale in Germany and used as food and feed. Single cell protein 63 Micrmrganisms are rich in protein. Microbial cells can contain as much protein as conventional foods. Bacteria can contain 60-696 (as a 96 of dry weight) protein whereas fungi and algae contain about 40%. In addition, microbial cells can be a rich source of fibre, unsaturated fats, minerals and vitamins. They are low in saturated fats and sodium. protein from inorganic nitrogen autotrophs/ heterotrophs 8aV88 agricultural spa- rapid gmwth Micrmrganisms create protein. Like plants, many micmrganisms can use inorganic nitrogen and can thus be used as an alternative to plants to create protein. In SB processes inorganic nitrogen is usually supplied as ammonia (or as ammonium salts), which is readily available and is renewable, as it can be manufactured from atmospheric nitrogen and can be recycled through the nitrogen cycle. Microorganisms can use alternative carbon sources. Algae are autotrophs using atmospheric COZ (think of them as plants growing in water instead of soil). They compete with plank for COZ but there is not a shortage of COz in the atmosphere and it is renewable by recycling through the carbon cycle. Other micm-organisms are heterotrophs (organisms which use organic sources of carbon) and can use a wide range of organic carbon sources. These can be materials unsuitable as food sources for animals (for example methanol). Others are waste products from industries or agriculture and have limited uses and can be a problem to dispose of by other means. SCP processes are efficient on space. SCP production plants can be built on land unsuitable for agriculture and so need not compete for space with conventional sources of protein. Also they are much more efficient in terms of amount of protein produced per unit area (figures are quoted later for some processes). Micro-organisms grow rapidly. Micrmrganisms grow much more rapidly than plants or animals. Bacteria can grow with mean generation times (doubling times) as short as 20-30 minutes. The mean generation times of unicellular algae and fungi are about 1-3 hours, whereas those of multicellular algae and fungi may be longer. This means that micmrganisms have the potential to produce protein far more rapidly than is possible by rearing plants or animals. By completing the following calculation you will be able to demonstrate the n amazing potential for micmrganism to rapidly produce protein for food. In batch culture, when growth is exponential, the number or organisms produced from one organism is given by 2”, where n is the number of generations. So after one qneration there are 2’ (ie 21, after two generations 2* (ie 4) and after three generations 2 (ie 8) and so on. Starting from a single bacterial cell with a mean generation time (doubling time) of 1 hour, and assuming exponential growth throughout, how many %anisms would you have after 48 hours? As the dry weight of a bacterial cell is about 10- g, what would the dry weight of these cells be? Assuming these cells to be 50% protein, how much protein would there be? Assuming you are an average person, you require about 70 g of protein in your diet per day. How long would this protein last you? (Do not cheat, try the calculation before reading on). Now repeat the calculation to find out how much protein you would have after 721. After 48 hours there would be 2& or 2.8 x 1014 cells. This represents 2.8 x 1014 x lo-’’ = 2.8 x 104g dry weight cells. 64 Chapter 4 This represents 2.8 x 1d x 50% = 1.4 x ldg protein. This repments 1.4 x 1$/70 = 200 days worth of protein for one person. After 72 hours there would be 2n or 4.7 x Id' cells. This represents 4.7 x Id' x lo-'' = 4.7 x 1O"g dry weight cells. ms represents 4.7 x IO" x 50% = 2.35 x 1o"g protein. This represents 2.35 x 10"/70 = ca. 3 x lo9 meals, or enough protein to feed the entire population of China for 3 days! You have now demonstrated the capability of exponentially increasing microbial populations to rapidly produce protein. However, although such outputs of protein are possible in theory, they cannot be achieved in practice, since exponential growth cannot be maintained for such periods because bioreactors are limited in size. 4.3.2 The disadvantages of micro-organisms as a protein source Let us consider some of the disadvantages of micro-organisms compared to conventional sources of protein. amino acid profiles digestibility importance of organoleptic properties edverse effects cost =w uric acid Are they nutritious? Guidelines on the nutritional quality of s8 have been given by the Protein Advisory Group (PAG) of the United Nations, and are based on amino acid profiles and feeding trials in animals. While most microbial cells are rich in protein, many do not contain sufficient quantities of essential amino acids. For instance, algal and fungal cells tend to lack methionine. Microbial cells may not be as easy to digest as conventional protein sources, for instance algae have cellulose cell walls which must be broken up if the proteins within the cell are to be easily digested by humans. The requirements for food are more strict than those for feed. What are they like to eat? Humans are particular about the organoleptic properties of their food. Microbial cells may have little taste or smell, or even smell or taste unpleasantly to some people. The texture may not be the same as in conventional foods, particularly with unicellular organisms. These draw-backs can be ovexrome by adding a proportion of SCP to manufactured foods. However, even when SCP is incorporated into manufactured foods it may not have suitable characteristics such as stability, ability to bind water or fats, or ability to form gels, emulsions or foams. SCP for feed does not have to meet such strict requirements. What happens after you have eaten them? Even if a micro-organism is palatable it may not necessarily be acceptable to the human digestive system, and if eaten in quantity can produce indigestion, flatulence, nausea, vomiting or diarrhoea. As little as 15 g yeast cells per day can produce such effects in humans. How much do they cost? SCP must compete in price with conventional protein foods and feeds. In countries where protein foods are readily available they can be relatively cheap. It has not always been possible to produce SCP at competitive prices. Are they safe to eat? Micmrganisms which are pathogenic or toxic obviously can not be used as SCP sources. In addition most microbial cells have a higher content of nucleic acid, particularly RNA, than conventional foods. When such cells are digested by animals these nucleic acids are metabolised to uric acid. Unlike most other mammals, humans do not possess uricase, which oxidises uric acid to soluble allantoid for Single cell protein 6!5 excretion, and so uric acid can build up in the blood and may deposit as crystals in the pints, causing gout and arthritis. Thus, SCP used as food is usually processed to reduce the RNA content. PAG guidelines recommend that for humans the daily intake of nucleic add should not be more than 4g, of which not more that 2 g should be obtained from SCP. If SCP has a nucleic acid content of 1596, how much of that s8 could be safely ingested per day? If the SCP contains 50% protein, what proportion of the mommended human daily requirement (of 70 g protein) does this represent? Try to work these out for yourself before reading our answers. n Our calculation: 29 nucleic acid would be present in 2 x 15 g =13.3 g SCP At 50% protein this represents 13.3 x 100 = 6.65 g protein This corresponds to - x 100 = 9.5 76 of the daily requirement. 100 50 6.65 70 Which of the following factors supports the use of micro-organisms rather than higher plants for the production of protein food? 1) 2) 3) 4) 5) Plants are more difficult to digest than micro-organisms. Micro-organisms can be used to convert organic wastes into proteins. Micmrganisms grow more quickly than plants. Higher plants need CG as a carbon source. Micmrganisms can use inorganic nitrogen. Suggest ways of overcoming or bypassing the following disadvantages of SCP as food. 1) Unpalatability 2) Indigestibility 3) Poor amino acid profile 4) Toxicity We have seen that only certain micro-organisms that conform to nutritional and safety requirements are suitable for food or feed, and that food has more strict requirements than feed. In addition, for use as food, SCP should have a reduced nucleic acid content and should be palatable. Most often this means that its use is limited to processed foods, in which food technologists can produce acceptable tastes, smells and textures. 66 Chapter 4 4.4 Substrates for SCP production For a micrmrganism to grow it must be supplied with all the nutrients required for ceII material and energy production. The physiological types of organism used in SCP production and their corresponding substrates are shown in Table 4.1. Photosynthetic bacteria utilise CQ from the atmosphere and nitrate in inorganic salts or natural ground water media. Algae are similar, growing on nitrate, ammonia or ammonium salt as nitrogen source. Some can also be grown as heterotrophs, in the dark, using sugars as sources of carbon and energy. Heterotrophic bacteria and fin@ for SCP are grown on a variety of organic substrates, serving as both carbon and energy sources. Some organisms have additional requirement for growth factors, such as vitamins. For yeast, the substrate is in the form of sugars, as yeast cells cannot break down polysaccharides, whereas filamentous fungi may in addition be able to use starch (by secreting amylases), pectin (by secreting pectinases) and cellulosic material (by secreting cellulases and hemicellulases). Waste containing cellulosic material is in solid rather than in liquid form. Processes have also been developed with yeasts growing on n-paraffins or ethanol, and with bacteria growing on methanol. Inorganic nitrogen is supplied in such processes as ammonia, or as ammonium salt. pectinases cellulases and hfli&lulmes Organisms Physiological Carbon Energy Nltrogen type source source source Blue-green Autotroph Atmospheric bacteria c02 Algae Fungi Autotroph Atmospheric Heterotroph Carbohydrate Heterotroph Carbohydrate con (sugars) (sugars, starch, pectin, cellulosics) Hydrocarbons and derivatives (n-paraff ins, ethanol) Bacteria Heterotroph Hydrocarbon (methylotroph) derivatives (methanol) Sunlight Sunlight As carbon source As carbon source As carbon source As carbon source N03- NH3, NH4+ N03- NH3, NH4+ N03' NH3, NH4+ NH3, NH4+ NH3, NH4+ Table 4.1 Organisms and substrates in SCP production Single cell protein 67 CN ratio ~lii-aubstrate fermentations renewable In a culture medium for the growth of heterotrophs what do you think the n carbon : nitrogen (CN) ratio should be? 1:l 1O:l 1oo:l 1Ooo:l The CN ratio should be about 101. The organic carbon in the medium provides both a source of energy and a source of carbon. Cells contain more carbon that nitrogen. The correct ratio of CN is about lO:l, although this differs slightly between organisms. This means that a medium containing 3% w/v sugar should be supplied with ammonia at about 0.3% w/v. At a CN ratio of 1:l most of the ammonia would not be incorporated into cells (it is present in excess) and would be wasted. At CN ratios more than 10:l the ammonia would be completely used up before all the sugar, ducing the biomass output and wasting the sugar. The cost of substrates used in SCF' production may represent 40-7596 of the total production cost. Ammonia contributes 515% of the substrate cost but the major portion is the carbon source. Atmospheric COZ is free, but costly energy is needed for agitation to dissolve it into dense algal cultures. Wastes from agriculture and industry can be plentiful and relatively cheap, but may still re resent 20-3096 of the total roduction pretreatment before they can be used in solid-substrate fermentations. Industrial wastes in the form of effluents can have high levels of BOD (Biological Oxygen Demand), which means they could cause pollution if disposed of in water without treatment Using them as substrates for XP production can reduce the BOD by as much as 70-801 and so save on treatment costs. Such agricultural and industrial wastes are derived from biomass (plant material) which is renewable and likely to remain plentiful and relatively cheap. Hydrocarbons and their derivatives can represent from 3G7096 of total production costs. They are derived from oil or natural gas which are non-renewable, will not remain as plentiful as at present and will become increasingly expensive. They also have alternative uses as fuels and petrochemicals and their availability is often influenced by political issues. costs. Solid agricultural wastes, especially ce 2 ulosic ones, may also n ee8 expensive 4.5 Micro-organisms for SCP production The physiological gruups of organisms used in XP production have been given in Table 4.1. We have exarmned ' the characteristics an organism should and should not have in order to be suitable as food or feed in Section 4.3. When selecting an organism for a particular production process, factors relating to growth of the organism also need to be considered. Listed below are characteristics in culture of an organism you are intending to use in an s8 process you are developing. Consider whether each characbistic is an advantage or disadvantage to you. Tick the appm riate box, or if you think tick both boxes. n the characteristic is an advantage on the one hand but a B 'sadvantage on the other, 68 Chapter 4 advantage disadvantage i) Low growth rate. ii) High biomass yield coefficient. iii) Filamentous growth. iv) Tolerance to broad range of temperatures. v) Tolerance to broad range of pH. vi) High spontaneous mutation rate. vii) Low aeration requirement. Output i) Disadvantage. High growth rate is needed for high output (weight of biomass produced per unit of time). The only advantage could be that as the RNA content of cells is generally proportional to the growth rate, growth at low growth rate could result in a product with lower nucleic acid content. Advantage. The biomass yield coefficient (weight of cells produced per unit of substrate consumed) should be high in order to give a high output. It also ensures efficient utilisation of the (expensive) substrate. wall growth iii) AdvantageDisadvantage. Compared to unicellularorganisms, filamentous ones are easier (and cheaper) to recover from fermentation media (by sieving or rotary vacuum filtration) and have a more fibrous texture. However, dense broths of filamentous organisms can be difficult to aerate and wall growth can cause problems such as clogging of pipes and valves. Advantage. Temperature increases can occur during fermentations, as growth processes are exothermic. The ability of an organism to tolerate raised temperature would reduce the need for Cooling. The ability of an qanism to grow at ambient temperatures also overcomes the need for heating and cooling. The broader the temperature range tolerated, the less the need for temperature control. Advantage. The pH of a medium tends to change during fermentation. Most often media are buffered, and the fermentor is fitted with pH control. However, the ability to tolerate a wide range of pH can overcome the need for pH control. Fungi generally grow at lower pH than bacteria. Use can sometimes be made of this by operating fungal processes at very low pH, preventing bacterial growth. This means that an aseptic process (using sterilising procedures to maintain a pure culture) will be less prone to contamination if aseptic procedures fail. In some circumstances non-aseptic (non-sterile) processes can be operated, saving sterilisation costs. Disadvantage. Organisms for SCP production quire a high degree of genetic stability. We have been considering the characteristics an organism must have for s8 production. These characteristics are under genetic control and any mutation ii) iv) v) non-aseptic processes vi) Single cell protein 69 could lead to an undesirable change in them. This is particularly important in continuous culture, which is often used for SCP production. Here the long growth period (in principle indefinite but in practice several weeks) can be long enough for mutants to arise, compete with the parent organism and predominate. In batch culture there is not enough time for this to occur. Advantage. For most SCP fermentation processes the running costs (costs of operating the fermentation unit) is W20% of the total production cost. Aeration costs contribute 3040% to running costs. In other words aeration costs can be as much as 12% of the production costs of the SCP. The lower the aeration costs the better. Production costs for various s8 processes are discussed in more detail later on. vii) You are now familiar with the mapr characteristics of organisms that are useful for SCP production, and the types of substrates on which they can be grown. We are now going to consider in detail the processes that have been developed. Some of these processes have been developed only as far as the pilot scale, and have not reached commercial operation. Others have reached full production scale but have subsequently failed, for a variety of reasons. These have been included as well as the successes, as they show you the variety in the technology of SCP production, and also show how economic and political factors influence the success and failure of processes. These processes might also become useful and economic some time in the future. Emphasis will be put on the technology involved in the fermentation and down-stream processing of each process. 4.6 SCP from carbon dioxide 4.6.1 Spirulina Blue-green bacteria (cyanobacteria) of the genus SpiruZina have been produced as SCF’ in Mexico, using natural bicarbonaterich ground-water (into which atmospheric CG readily dissolves). A flow diagram of the process is given in Figure 4.2. The single 10 hectare (1 ha = l0,OOO m’ )open lagoon is about 0.6 m in depth and unmixed. The system is operated as a batch culture or as a semi-continuous culture (in which a proportion of the medium is removed and replaced by fresh medium intermittently). Nitrate is added as a nitrogen source, and other minerals are present in the water. The long filaments are raked mechanically from the pond onto screens (sieves), where water is drained and either recycled or disposed of. The biomass is then dewatered by rotary vacuum filtration, dried by vacuum drying, then dried in a drum drier and ground to a powder (to make the product more appealing). The product contains 56% protein and is sold as food. The plant operates with an output of 10 g dry weight per square metre per day. n What is the output of the 10 hectare lagoon per year? (Note 1 hectare = l0,OOO m 1. 10 g m-2 equates to 10 x lo4 = leg ha? The daily output is thus: 10’ kg ha:’ day-’. The lOha lagoon produces Id kg ha:’ day-’. This equates to Id x 365 = 3.6 x Id kg year-’. The output of the lagoon is thus 360 tonnes year -’. open lagoon system 70 Chapter 4 Figure 4.2 The production of Spirulina maxima for SCP 4.6.2 Algae Eukaryotic algae (Chlorophyceae) of the genera Chefla and Scenedesmus have been used for SCP production. Several types of cultivation systems have been considered, depending on the substrate used and whether the SCP is intended for use as food or feed. Single cell protein 71 a2 enrichmenV phototroophe centrifugation flocarlation drum drying For use as food or feed, algae are grown in pure or mixed culture in a mineral salts medium containing or NQ- and supplied with air or gaseous COz (air contains only0.0396 Ca so Co2 enrichment is required for high output). A flow diagram of such a process is essentially the same as that shown in Figure 4.2. Open systems have been developed with organisms growing as continuous cultures in open lagoons or circulation ditches, similar in design to oxidation ponds and ditches used in sewage treatment. The ponds can be lined with clay, concrete, brick or plastic sheeting and axe 20-50 an in depth. Wng can be mechanical, using motor driven paddles, or can be manual, and it is necessary to prevent sedimentation of cells and uneven exposure to sunlight. Capital costs (excluding land costs) are lower than with other SB systems, at $(US)20,000-30,000 per hectare (1989 prices). The organisms grow as phototmphs using sunlight as an energy source and atmospheric COz as carbon source. Such systems are relatively simple (low-tech) but, as they are open, they are liable to contamination by wild algae and bacteria. Heterotmphic bacteria can grow in the ponds using organic materials released into the water by the algal cells. Write down three reasons why the requirement for sunlight limits the use of algal n ponds in SCP production? Firstly, the lagoons need to be open to allow the sunlight to penetrate. This means that pond culture is non-aseptic and contamination usually limits the use of SB to feed, where a higher degree of contamination is acceptable. The micro-biological standards of SCP as food and feed have been defined by the PAG and are based on comparisons with conventional foods and feeds. Secondly, the technology can only be applied between latitudes 35O North and South where there is sufficient sunlight to give high outputs. Thirdly, in dense algal cultures sunlight does not penetrate more than 40-50 an, so lagoons are limited to such depths. This limits output, ie you cannot produce more algae by making the lagoon deeper. Cell recovery presents a problem in algal culture, as centrifugation (the most effective method) can be prohibitively expensive (Table 4.2). Methods such as flocculation with calcium hydroxide and sedimentation can be employed for feed purposes. Waste water is recycled and recovered cells are dried preferably by drum dryins, which breaks up cell walls making the product easier to digest by humans. The product contains 40-5096 protein and 44% nucleic acid, and has been produced in relatively small quantities at a selling rice $4-10 kg-’ (1990 prices). This compares to soya protein concentrate (70% protein P and milk powder (36% protein) at about $3 kg-’. Process Cost (US $ per Concentration thousand m3 of wtture) factor Centriiugation 380 40-1 00 Flocculation/Flotation 340 85 Flocculatlon/Sedlmentation 320 50 Table 4.2 Costs of algal harvesting processes (1 989 prices) 72 Chapter 4 In algal pond cultures cell concentrations of 2 g dry wt 1-1 are possible, corresponding to outputs of 15 g dry wt per square metre per day. What would the protein output of such a system be (as kg protein per hedare per year), assuming the dry algal biomass contains 45% protein? (Use a piece of paper to do this calculation and then compare it with ours). n 15 g dry wt m-2d-' corresponds to 15 x lo4 = 1.5 x lo5 g ha-'d-'. The annual output is thus 1.5 x Id x 365 = 5.5 x 107g hi' yf' or 5.5 x lo4 kg ha-'yf'. At 45% protein this corresponds to 5.5 x lo4 x 45/100 = 2.5 x lo4 kg protein ha-' yr-'. With increased C@ concentrations, by the injection of C@ gas into ponds, outputs of 3 x lo4 kg dry protein per hectare per year are possible. This compares very favourable with conventional sources of protein (Table 4-31, note the low yields of meat and milk. I Sources kg proteln per hectare per year I Algal ponds 30,000 Farmed fish Potatoes Rice Peanuts Wheat Milk Meat 1,000 800 600 450 360 120 80 Table 4.3 Output d protein from algal ponds and conventional sources tubular loop A novel fermentation system, a tubular loop reactor (Figuxv 4.31, has been developed at laboratory scale which is capable of producing algal cultures with densities of 20 g dry wt 1-'. This system converts 18% of incident solar energy, far in excess of the 7% in algal ponds and 1-2% in agriculture. Such a s stem 'ves a theoretical output in European climates of 1oO,OOO-15O,ooO kg protein ha' year -rand could be operated in arid mons as the water is conserved and can be recycled. Single cell protein 73 Figure 4.3 Tubular-loop reactor for the production of algal biomass The photosynthetic production of Chlorella sp can be written as: light COZ + H20 + NH3 > Biomass + 02 Development work is being carried out growing algae for feed on muniapal effluent8 and animal slumes (from intensive animal farms). This is carried out in high-rate algal ponds (shallow aerated lagoons operated at high dilution rates). Aerobic bacberh oxidise organic materials in the effluents, producing COZ which is used by the algae growing as photoautotrophs (using COZ and sunlight). The algae in turn produce 0, which further stimulates the aerobic bacteria. Such systems are able to produce feed an the one hand and to reduce BOD, nitrate and phosphate (ie pollutants) from effluents on the other hand. In Japan CWmZla spp has been produced for food in continuous aseptic systems in conventional bioreactoxs. The organisms are grown in the dark as heterotrophs udng sucrose (in the form of molasses) or glumse as carbon and energy source. Producth has been 2,ooCr3,000 tonnes per year at a selling price of $(US)10-22 kg-' (1990 PriCJes). This product is sold as a high-value health food. high-rate algal pond* 74 Chapter 4 Choose the correct completion to the following statement. In the production of Chlorella spp for s8 from molasses, bioreactors rather than lagoons are used because: 1) molasses-based medium is darkly coloured and would not allow sunlight to 2) the COZ required for growth can be more efficiently dissolved in fermentors; 3) lagoon systems containing molasses-based medium would become heavily 4) bioreactors give higher biomass yields than lagoons. penetrate lagoons; contaminated by bacteria; 4.7 SCP from carbohydrates Carbohydrates derived from plant biomass are plentiful and are renewable. They thus form an excellent substrate for s8 production by heterotrophs. Such carbohydrates can be in the form of sugars, which are readily used by microbial cells, or starch, pectin and cellulosic material, which require hydrolysis to sugars before they are available for cell uptake and use. Hydrolysis of starch can be carried out by chemical or enzymatic treatment to produce sugars for a fermentation medium, or by enzymes produced in situ by an amylaseproducing organism growing in or on the starch substrate. Cellulosic material is solid and is normally used in solid-substrate fermentations. The estimated quantities of such materials available for SCP production (or other fermentation processes) are given in Table 4.4. Note the importance of grainderived products. Single cell protein 75 Lpe SOUrCe Chantlty (x 10' tonner per ywr) Sugars: Starch: :ellulosics: Molasses Whey Sulphite waste liquor Glucose (from starch) Wastes from fruit processing Wastes from vegetable processing Bagasse (sugar cane fibre) Wheat bran Wheat straw Rice straw Corn (maize) wastes Urban refuse (paperkardboard) Woodnorestry wastes 9.3 1.5 (USA) 12-2 1 06 58 864 599 1 93 152 61 Table 4.4 Approximate world (or USA) availability of carbohydrate substrates 4.7.1 Sugars Yeasts are the product of many of the SCP processes based on sugars. This is because yeasts can use many different sugars, they have been used traditionally in foods and they have most of the desirable characteristics for SCP. Each process differs sli tly systems are similar. The experience you gained in earlier sections should enable you to answer the following SAQ. yeasts according to the nature of the substrate and the organism used, but in principle afthe 76 Chapter 4 The diagram below is a flow diagram of an SB process from carbohydrate, with the letters A-K representing various operations, inputs and outputs. Using the list given, fit in appropriate operations, inputs or outputs at points A-K to complete the flow diagram. mobses Molasses is a by-product of sugar rdining, mostly from beet or cane, and contains 35-50% suame and small quantities of nitrogen. It is used as sweeteners in the food industry and as a fermentation medium for the production of bakers yeast, ethanol and other products. For SCP production, the molasses is diluted to 44% sucrose, supplemented with phosphate and sterilised by continuous heat sterilisation. Continuous processes are run in aerated fermentors with ammonia addition, producing food-grade Candida utilis and Sacchmomyces cerm*siae. Systems operate at dilution rates 0.2-0.3 h-’ at pH 3.545 at 25-35°C. Yeasts are recovered by centrifugation, washed, dried by drum or spray drying and packaged. The product contains about 45% protein and is used as a high protein food supplement, particularly in Taiwan, South Africa and the former USSR. -tim~~us Culture Single cell protein n Presented below are four flow diagrams representing processes for washing the concentrated yeast suspension (the suspension concentrated from the fermentation medium by centrifugation). Which of these processes do you think would be suitable for washing the yeast suspension? rI Methods 1 and 4 are appropriate methods of washing yeast cells. Yeast cells can be concentrated by centrifugation, so method 1 achieves washing. Method 4 makes use of the fact that concentrated yeast cells can be dewatered using a rotary vacuum filter which can incorporate washing. In order to filter yeast cells the filter would first have to be coated (pre-coated) with a layer of material such as starch granules, to decrease the effective size of the filter. Method 2 does not achieve washing, as medium components would be dried with the yeast and not removed in the water. Method 3 is not suitable for concentrating large volumes of yeast cells as ultrafilters would become blocked by dense yeast suspension. Whey is the effluent from cheese manufacture, and contains about 5% lactose and about 1% protein. About half of the global production is used as a feed supplement but the rest is unused. The BOD can be as high as 70,000 mg 02 1-’, which rrquires reduction by biological treatment (which is costly) prior to disposal. For SCP processes the valuable protein from the whey is first recovered by precipitation or ultrafiltration, and the deproteinised whey used in processes such as described for molasses. The yeasts used rotary vacuum filter whey depmtination 78 Chapter 4 are food-grade organisms capable of using lactose as a carbon source, namely Saccharomyces lactis, Candida utilis and Kluyumyces (previously Saccharomyces) jkzgilis. Feed-grade SCP is often prepared by spray drying the complete fermentation broth. A flow diagram of a process producing SB from whey is shown in Figure 4.4. Figure 4.4 The production of whey protein and K/uperumyces /actis from whey. sulphite waste liquor Sulphite waste liquor is a waste product of the sulphite wood pulping process. This process has now been replaced in many countries and the availability of sulphite waste liquor is now less than that indicated in Table 4.4. The liquor contains about 20% w/v sugars, in the form of both hexoses (6carbon sugars) and pentoses (5-carbon sugars), and 6% w/v acetic acid and has a BOD of up to 50,000 mg G 1". Processes have been operated growing Candida utilis for food or feed but the most effective has produced feed from Paecibmyces azrioti (the Pekilo process). This organism more effectively uses pentose sugars and the acetic acid present in the substrate compared to With spp, and so gives higher output and greater BOD reduction. The organism grows as short filaments (making cell recovery less costly) and has a pmtein content of 5540%. Any similar carbohydrate source from food processing (such as coconut milk or effluents from fruit canning) can be treated in this way, and there are many examples of development studies producing s8 from them. Single cell protein 79 n The conversion of carbohydrate to biomass can be represented as follows: Studies show that the production of lkg dry biomass requires 2.0 kg sugar, 0.7 kg oxygen, 0.1 kg ammonia, with the liberation of 12,300 k Joules heat. A typical continuous fermentation operates at a dilution rate (D) = 0.2 h-’, wi&h sugar concentration of 3% (w/v) in the incoming medium. With a fermentor of50 m capacity and 90% utilisation of carbohydrate [ie 0.3% (w/v) sugar in the outgoing medium] what would be: 1) 2) 3) 4) 5) the biomass concentration (quantity biomass m-3); the productivity (quantity biomass m-3h”); the output rate (quantity biomass from the system h-’); the minimum oxygen-transfer rate (OTR) (quantity 02 transferred m-3h”); the heat evolution rate (quantity of heat m-3h”)? Again we urge you to do the calculations on a piece of paper before checking your answer with ours. Biomass concentration. 1 kg biomass is produced from 2 kg sugar, so the growth yield is 1/2 = 0.5 The incoming medium contains 3% sugar (30 g 1-’ or 30 kg ma) and the outgoing medium 0.3% sugar (3 g 1-I or 3 kg ma). Sugar utilisation is thus 3@3 = 27 kg m-3. With a yield factor of 05 the biomass concentration would be 27 x 0.5 = 135 kg m”. Productivity. The continuous system operates such that the biomass concentration remains constant (specific growth rate = dilution rate). Productivity is therefore given by the biomass concentration x the dilution rate. The dilution rate is 0.2 h-’, so the effect of dilution or growth is 135 x 0.2 = 2.7 kg m-%-’. The productivity is therefore 2.7 kg m” h”. Output. Output = productivity x reactor volume = 2.7 x 50 = 135 kg h”. Minimum OTR. The production of 1 kg biomass requires 0.7 kg 02. The productivity is 2.7 kg m-3h-1 (from 2 above). 2.7 kg biomass thus requirps 2.7 x 0.7 = 1.89 kg 02. The minimum OTR is thus 1.89 kg 02 m-3h-I. Heat Evolution. The production of lkg biomass liberates 12,300 k Jdes. The productivity is 2.7 kg m-3h”. 2.7 kg biomass involves the liberation of 2.7 x 12,3000 = 33,210 k Joules. The heat evolution is therefore 33,210 k Joules m-3h”. If you were unable to do these calculations, we would suggest that you read the BIOTOL text entitled ’In mho Cultivation of Micm-organisms‘ or other suitable texts of the fundamental principles of cell cultivation. 80 Chapter 4 n 1) 2) 3) 4) For SCP production from carbohydrates, aerobic rather than anaerobic growth is preferred, as aerobic growth leads to higher biomass yield coefficients (kg biomass produced per kg substrate used). Choose the correct reason for such higher yield coefficient from the following. Anaerobic utilisation of carbohydrates is strongly inhibited by the end product ethanol, giving low yield coefficients compared to aerobic utilisation. Aerobic growth utilises the glycolic pathway (for converting carbohydrate to pyruvate) yielding more energy than anaerobic growth, which utilises the pentose phosphate pathway, thus giving higher yield coefficients. Aerobic growth (converting pyruvate to Co2 and HzO) yields more energy (ATP) than anaerobic growth (converting pyriivate to ethanol and CG), thus giving higher yield coefficients. Aeration required for aerobic growth removes ethanol (produced by yeast growth) from the medium, keeping ethanol levels below inhibitory concentrations, thus giving a higher yield. Response: Incorrect - Ethanol can inhibit the growth of yeast at high concentrations (about 7% v/v). This could lead to lower productivity and output, and lead to higher residual sugar concentrations (lower degree of conversion). However this would not influence yield coefficients, which relate to the amount of substrate actually Used. Incorrect - The two pathways are alternative pathways for converting carbohydrate to pyruvate, and are found in various organisms. They operate in aerobic or anaerobic conditions. Correct - Aerobic growth is more energy efficient (produces more ATP) than anaerobic growth, resulting in higher yield coefficients. Incorrect - Aerated cultures do not produce alcohol. Even if they did, point 1) above applies, and aeration would not necessarily reduce ethanol levels significantly. If you were unable to work out the answer to this it probably means you need to refresh knowledge of cell metabolism. BIOTOL texts on cell metabolism or other good introductory biochemistry text should prove adequate. Two types of bioreactors have been used to produce s(zp from sugar sources, namely stid/baffled type ayd the air-lift with draught tube type (Figure 43, with various capacities up to 400 m . Single cell protein 81 Figure 4.5 Fermentor types used in the production of SCP from sugar sources, (a) Stirred baffled system, (b) and (c) air lift systems myco-pmtein The patented 'My-protein' process, operating in the UK involves growing the filamentous fungus fusarium gruminaarum on food-grade glucose syrup, produced by the enzymatic hydrolysis of wheat starch, a byprodud of wheat gluten production. The medium is composed of diluted glucose syrup, ammonia, mineral salts, choline and biotin. Ammonia gas is added to continuous aseptic systems via the air inlet and serves as a nitrogen source and to maintain pH at 9.0. The system used to develop the process was an air-lift fermentor of 36 m3 capacity. For full-sale production a pressurecycle bioreactor of la r volume is used. This technology was developed for SCP production from methanol T see Section 4.9) and was acquired under licence for the Myco-protein process. Operation is at dilution rates of about 0.2 h-' with dry weights of 15-17 g I-'. These conditions are selected partly to enhance the filamentous (fibrous) structure of the biomass.:After concentration from the medium, by separation in a cyclone (whirlpool device), the cell suspension is heated to 64°C for about 20 minutes, during which time heat-stable RNAses degrade cellular RNA, the components of which are released from the cells. This reduces the RNA content from 10% to 2%. The heating process in addition inactivates proteases, which makes the produce more stable during storage. Biomass recovery is by filtration (dewatering) by horizontal belt vacuum filtration. The recovered biomass is usually blast chilled for storage by deep freezing, although it can be dried for storage. The product is given the trade ME Myceprotein. This product contains 45% protein on a dry weight basis. reduction Of RNA content Use the description given above to draw a flow diagram to show the production of Myco-protein. 82 Chapter 4 changing market circumstances diversification of product nutritional quality safety checks When the Myco-protein project was planned in the 1960s it was assumed the product would form a valuable protein source, needed to meet expected protein shortages. By the early 198Os, when the project had reached production, the protein shortage had not developed (at least in Europe). However, dietary habits had changed, and there was a move away from meat consumption to other 'healthier' foods. There was also a move towards 'convenience' foods such as deep-frozen prepared meals. It was to these markets that Myco-protein was directed. The dewatered fungal biomass has a natural fibrous texture (being composed of fungal hyphae). The hyphae can be partially aligned by rolling the biomass between rollers. If the biomass is premixed with egg albumin, rolled, then steam-heated to set the albumin, a material with meat-like texture is produced. By altering the processing conditions, material with textures of various meats (such as beef, chicken or fish) can be produced. The product itself does not have a strong flavour and so will take up the flavour of dishes to which it is added. Myco-protein was successfully tested for nutritional quality and safety in feeding trials in animals over a ten-year period. This included trials on human volunteers, carxied out in the UK and USA. In 1985 the product was approved by the Ministry of Agriculture Fisheries and Food for sale as food in the UK. The product is marketed under the tradename Quorn, and is promoted on the basis of having satisfymg meat-like texture with high protein and dietary fibre content while being without animal fat, low in calories and suitable for vegetarians. The nutritional quality of Quorn, compared to conventional protein sources is given in the Tables 45 and 4.6. Cheddar Raw Raw Stew Fresh Raw Raw Worn cheese chicken lean steak flrh pork beef beef (cod) (leg) =usage Protein (Yo wh) Dietary fibre (Yo wh) Fats (% w/w) Ratio polyunsaturated fatty acids: saturated fatty acid Cholesterol (mg/lOO g) Energy (k Joules/l OOg) 26.0 20.5 20.3 20.2 17.4 16.6 9.6 12.2 0 0 0 0 0 0 0 5.1 33.5 4.3 4.6 10.6 0.7 22.5 24.1 2.9 0.2 0.5 0.1 0.1 2.2 0.2 0.1 2.5 70 69 59 65 50 72 40 0 1665 496 504 721 311 1103 1226 328 Table 4.5 The nutritional quality of Quorn and some conventional foods Single cell protein 83 Isoleucine 4.0 4.7 4.4 5.0 4.6 4.5 Leucine 7.0 8.0 7.3 7.7 7.5 7.1 Lysine 5.5 6.3 6.8 8.8 9.0 7.4 Methionine 3.5 3.8 2.6 3.9 3.7 3.0 and cysteine and tyroslne Phenylalanine 6.0 8.2 7.9 8.3 8.0 8.0 Threonine 4.0 4.1 4.9 4.3 4.1 4.0 Tryptophan 1 .o 1.3 1.4 1.3 1.1 1.3 Valine 5.0 5.6 5.3 5.1 4.8 5.0 \ - MeeiDiet Avoraga Worn Beef Chicken Soya (WHO) DW(UK) Does Quorn have a high ratio of polyunsaturated to saturated fatty acids compared with meat products? How does its cholesterol and essential amino acid content compare with other sources? Quorn has a higher ratio of polyunsaturated to saturated fatty acids than the other foods listed. It contains no cholesterol. The amino acid profile is similar to that of other protein foods. It is slightly lower in methionine and cysteine content. 4.7.2 Starches Effluents from the processing of starchy vegetables such as potatoes, cassava, rice or corn (maize) have been the substrates for commercially operated SCP processes. The patented Symba process, developed in Sweden, is a two-stage continuous proms based on a symbiotic culture of the yeasts Endomycopsis fibuligena (which is amylolytic but of low value as SCP) and Candida ufilis. The effluent is supplemented with ammonia, sterilised and fed to the first bioreactor containing E. fibuligera. This organism secretes amylase which hydrolyses the starch. The broth feeds a second bioreactor inoculated with Candida utilis, which grows faster and predominates. The process reduces the BOD of the effluent by 90%, to a level of loo0 mg 02 per litre. Development work has been carried out on processes to produce feed from starch effluents in non-aseptic waste treatment systems (such as oxidation ditches and aeration ponds) using amylolytic filamentous fungi. These often belonging to the genera AspergzZlus and lUtiqms. In such processes, the levels of contaminating bacteria are depressed by operating the systems at low pH. n yrnbiotic culturn 84 Chapter 4 Such fungi have also been used in development studies on inmasing the protein content of solid starch wastes (vegetable rind or peel) or conventional feed (grains) by using solid-substrate fermentations. This involves chopping up the substrate, if necessary, and steaming it for about 1 hour (for hydration, and pasteurisation to reduce numbers of competing mimrganisms). After steaming, the substrate is cooled and the moisture content is adjusted, if necessary, to 5@75%. Ammonium salt and other necessary minerals are added, followed by inoculation and incubation. This type of process can be carried out on a small scale with manual aeration (by 'turning the pile'). On a larger scale, inoculated material can be incubated in tanks or bins with mechanical raking and forced aeration, or laid out in rows and turned mechanically. Once fungal growth is complete (after 2448 h), the material can be fed directly to animals or dried for storage. In this way starchy material low in protein can be converted into a more balanced feed material containing 10-15% protein. The product is called 'protein-enriched feed'. This technology has been developed with a view to utilising cellulosic wastes, which are produced on a much larger scale. This is discussed below. fermentations f~~~in*nri~~ feed Draw a flow diagram to show the production of protein-enriched feed from solid starch wastes. SAQ 4.6 I 4.7.3 Cellulosics You have seen from Table 4.4 that cellulosic material is produced in vast quantities as a waste. If an economic process can be found to hydrolyse cellulosic material to its constituent sugars, by chemical and/or enzymic means, then vast quantities of sugar could be made available for fermentation processes, including SCP production. Despite long and hard research effort, success has been limited. However, projects have been developed to enhance protein content of cellulosic wastes by growing cellulolytic fungi (such as Trichoderma spp) on them in solid-substrate fermentations (using the technology described for solid starch wastes). In this way the low-protein substrate (for example wheat straw or rice straw) can be converted into protein-enriched feed, with a protein content of up of 15%. This can be used as feed for ruminant animals. However, for high conversion rates, cellulosic material requires expensive physical or chemical pretreatment. As yet no largescale commercial processes have been developed. What problems do you think might be caused by unwanted contaminating fungi n in solid substrate fermentation processes for protein-enriched feed? The cells of many fungi are toxic if eaten. In addition several fungi can produce toxins (called mycotoxins) which cause food poisoning to man and animals (for example Aspm@Zus @s produces a toxin called aflatoxin). Other fungi can produce spores which can cause respiratory disorders in man and animals (for example Asperigillus fumigatus). These problems are minimised by using a heavy inoculum and so reducing the incubation period. Processes have been developed in North America to pilot scale growing the bacterium Cellulomonas or the fungus Trichaderma on pre-treated milled cellulosic material in conventional fermentors, ie in a liquid medium. However, preparation costs are considerably higher than with solid-substrate fermentations. hydrdyaisto constituent swers Single cell protein 85 4.8 SCP from hydrocarbons and derivatives In countries fortunate enough to have oil and gas deposits, or having access to such deposits, hydrocarbons are available as SCP substrates. This is particularly true of non-industrialised countries in warm climates where hydrocarbons have limited uses as fuels or petrochemicals. 4.8.1 n-alkanes Processes have been developed to production scale growing yeasts, for feed, on purified Cl0 -Ca n-alkanes (such n-paraffins being liquid at normal ambient temperatures). A flow diagram of the process producing Candida Zipolytica is given in Figure 4.6. rdkanes Figure 4.6 The production of Candida /i,c?/ytica from n-alkanes You should now be completely familiar with this type of flow diagram as it is similar to that shown for the production of SCP from whey. However, the characteristics of the fermentation differ in several respects. n-Alkanes are only slightly soluble in water and form droplets in stirred aerated cultures. Direct contact with droplets is necessary if yeasts are to use the substrate, so steps are taken to maximise the surface:volume ratio of droplets. This is achieved by controlling agitation and also sometimes by the addition of surfactants (detergents) to produce droplets of 1-100 pm diameter. Compared to carbohydrates, hydrocarbons are more highly reduced, and so requim more oxygen for their oxidative degradation. To produce 1 kg biomass from n-alkanes requires 1-1.2 kg n-alkanes and 2.2 kg G. This requires high oxygen-transfer rates (OTRs) by (expensive) vigorous agitation and use of over-pressures (operating fermentors at positive pressures to enhance gas dissolution). This adds signiiicantly to the cost of the product. To produce high OTRs, processes have made use of fermentor designs other than conventional stirred or air-lift systems, such as loop and tubular loop fermentors, and various stirring devices incorporating airsparing (Figure 4.7). lowsoiubitity oxygen transfer rate 06 Chapter 4 Figure 4.7 fermentor types used in the production of SCP from nparaffins Single cell protein a7 heat generation Oxidation of n-alkanes is strongly exothermic, the production of 1 k biomass liberating 27,100 k Joules. Cooling costs to maintain temperatures at about 30 C are considerable. Compare a) the minimum OTR and b) heat evolution rate of a continuous fermentation system based on n-alkafles operating at a dilution rate of 0.2 h-’ and a biomass concentration of 135 kg m- , to a similar system based on carbohydrate. You may need to look back a few pages to get the relevant information concerning carbohydrate utilisation (Section 4.7). 6 The process illustrated in Figure 4.6 was developed to production scale with a capacity of 200,000 tonnes per year. This process, developed by British Petroleum, was one of several in Europe and Japan that, although fully developed, was never operated commercially. This was due to sharply inmased substrate costs in 1973 and political and social pressures against the use of petroleum-based substrates (possibly contaminated with carcinogenic or toxic compounds). Such systems do operate in the former USSR, producing Crmdida gudliermundii as feed. Various wastes available as carbon substrates for SCP productions are listed in column A. Mat& each waste with a suitable organism (perhaps more than one) from column B. For each organism select the most appropriate production system from column C. A B C i) Candida utilis subtrate costs i) Waste paper ii) Exhaust gas emissions ii) Cbioreiia regularis ii) Solid-substrate fermentation from a coal-fired power station iii) Molasses iii) Aspergillus niger iii) Aqueous culture in iv) Effluent from fruit and iv) Tricbcxletma vifide vegetable processing factory containing 0.5% sugar and 0.5% starch i) Aqueous culture in open ponds or lagoons bioreactors 4.8.2 Ethanol Ethanol can be produced relatively cheaply as a bulk petrochemical by the hydration to ethylene. A few processes operate in the USA using ethanol to produce food-grade Cundida utih, with capacity of about 7,000 tonnes per year. The technology is similar to that already described for SCP from sugars. On ethanol, the yield is about 0.7 kg dry wt per kg ethanol used. 4.8.3 Methanol Methanol can be produced relatively cheaply as a bulk chemical by the oxidation of methane. Several processes have been developed to produce feed-grade SCP using methanol as a substrate. We will now examine one such process in depth, to show how a process is developed, from inception to production scale, and how the many problems encountered can be tackled and overcome. 88 Chapter 4 4.9 The Pruteen process - a case study 4.9.1 The planning stage In order to fully appreciate the driving forces behind the development of SCP projects you must imagine you have been transported back in time to the 1960s. It was then that the projects we have been looking at, and the one we are now going to look at in more detail, were planned. The world’s population is booming and it is expected that a severe world protein shortage will develop before the year 2,000. European economies are recovering from the effects of war, and increasing prosperity is leading to increasing consumption of meat, which had been rationed throughout much of Europe during the war years and for some time after. Meat rationing is therefore still very much in peoples memories. Europe has never been self sufficient in animal feed, and relies on imports of soya from the USA and fish meal from South America. This, then, was the Scene in which SCP projects were planned in Europe. The need for alternative foods and feedstuffs was clear and, in the UK and elsewhere, oil and gas seemed a plentiful and cheap resource from which to produce them. No& Sea gas fields were being exploited and research had shown that ~tural gas or its derivatives could be used to produce SCP feed of superior protein content to conventional feedstuffs. The economics of such processes seemed very promising. In 197l the European prices for fish meal and soya meal were $200 and $100 per ton respectively. In 1973 oil price rises and a failure in the Peruvian fish meal supply pushed these prices up to $550 and $300 respectively. With such prices for the mapr feedstuffs it was considered that SCP feeds could be produced competitively. 4.9.2 Methane as a substrate In the late 1960s, Imperial Chemical Industries (ICI) in the UK were interested in developing an SCP process using abundant and cheap methane from newly developed sources in the North Sea. However, it soon became apparent that methane was unsuitable as a substrate for fermentation. historical PmPdv* Answer true or false to each of the following statements. Methane is unsuitable n as a carbon source for SCP production because: 1) micro-organisms have not been isolated that are capable of using itas a sole source of carbon and energy; at normal temperatures and pressures it has low solubility in water, limited productivity and output; it is potentially explosive when mixed with air; it is a gas, relatively difficult to store and transport; it is difficult to obtain in a pure form; it is toxic to micro-organisms at high concentration. 2) 3) 4) 5) 6) Single cell protein 89 False. The ability to utilise methane as a sole source of carbon and energy (methanotrophy) is found in several genera of bacteria, such as MethyZumonus, Pseudomonas and Methylmcus. True. Even at high overpressures the low solubility of methane limits productivity and output. True. The explosive nature of methane:& mixture requires careful operation of largescale systems and incurs high insurance prrmiums. True. Methane is expensive to liquefy and, as a gas, is more difficult to store and transport than a liquid. False. Methane, as a gas, is much easier to obtain in pure form from oil and gas deposits than, of example, n-alkanes. False. Methane is non-toxic to micro-organisms. biomass yields biochemical pathway methane monooxygenase Although from theoretical considerations biomass yields from methane could be as high as 1.4, in laboratory-scale cultures values of about 1 .O were obtained, and in larger scale systems values were around 0.3-0.6. Methane fermentation also incurs high aeration and cooling costs. The biochemistry of methanotrophic bacteria was studied to show how energy, reducing equivalents and C3 skeletons were produced from the CI substrate. Methane is oxidised to methanol then formaldehyde. Formaldehyde can be assimilated, by various mechanisms, with Cs sugars, to form 6 sugars. The G sugars are either assimilated, recycled to CS sugars or converted to C3 skeletons such as pyruvate or glyceraldehyde (see Figure 4.9, section 4.9.6). The lower than expected yields can be explained by the nature of methane oxidation to methanol in these bacteria. This reaction, catalysed by methane mono-oxygenase, is a net consumer of reducing equivalents (NADH), which would otherwise be directed to ATP generation and biosynthesis. In simple terms the oxidation of methane to methanol consumes energy, lowering the yield. From the above discussion, what do you think could be used instead of methane n as a substrate? Formaldehyde is highly toxic and reactive and not suitable as a growth substrate. Methanol is more soluble in water than methane and would seem a suitable substrate. 4.9.3 Stage one - methanol as a substrate Methane can be chemically oxidised to methanol relatively cheaply. The heat liberated during such oxidation is higher grade than that liberated by biological conversion, and is thus easier to recover and use elsewhere. It was expected that using methanol as a fermentation substrate would, having by-passed the energy-inefficient methane oxidation step, lead to higher yield and lower the oxygen and cooling requirement. Methanol has the added advantage of being very much more soluble in water and easier to handle than methane. 90 Chapter 4 At that time few micro-organisms capable of using methanol as sole source of carbon and energy (methylotrophs) had been isolated, and so steps were taken to isolate such organisms from samples of soil, water and vegetation. me*YlotroPhs Protocols for isolating methylotrophic micrmrganisms by enrichment from soil, with a view to SCP production, are given below. Choose the most appropriate one. n Add 10 g soil to looml liquid Medium A. Incubate with shaking in darkness for 24h at 22°C. Spread 10 x 0.1 ml volumes on agar plates of Medium A and incubate at 22°C until colonies appear. Add 10 g soil to lOOml buffered saline. Shake vigorously for 10 minutes. Spread 10 x 0.1 ml volumes on agar plates of Medium A and incubate at 22°C until colonies appear. Add 10 g soil to 100 ml liquid Medium B. Incubate with shaking in darkness for seven days at 22°C. Spread 10 x 0.1 ml volumes on agar plates of Medium B and incubate at 22°C until colonies appear. Add 10 g soil to 100 ml liquid Medium C. Incubate with shaking in darkness for seven days at 22°C. Spread 10 x 0.1 ml volumes on agar plates of Medium C and incubate at 22°C until colonies appear. Medium A: 0.5% w/v (NH4hSO4; 0.4% w/v NaH2PO4; 0.08% w/v MgsO 4.7H20; 0.01% w/v FeSO4; 20.0% v/v methanol; pH 7.0 Medium B: 0.5% w/v (NH4hSO4; 0.4%W/V NaH2 Po4; 0.08% W/V Mg SO4.7H20; 0.01 w/v Fe SO4; 1.0% methanol v/v; pH 7.0 Medium C: 0.5% w/v (Nab ,504; 0.4% w/v NaH2 PO4; 0.3% v/v methanol; 0.3%w/v yeast extract; pH 7.0 All media contain purified agar (solidifjmg agent) lacking organic nutrients Response: Not appropriate. Medium A contains 20% methanol which is likely to be inhibitory or lethal to most organisms. 24 hours is also a short time to allow for enrichment to occur. Not appropriate. No enrichment occurs here, so the method is unlikely to pick up the very small numbers of organisms likely to be present in the soil. Medium A is also used, so response 1 also applies. Appropriate. The medium contains a suitable methanol concentration and allows a suitable time for enrichment to occur. Medium B is elective for methylotrophs, that is, it selectively stimulates the growth of this type of organism (in dark conditions). This is the basis of the process of enrichment, in which the small numbers of a particular type of organism are stimulated to incrrease to a point where isolation is easier. Single cell protein 91 4) Not appropriate. Medium C would allow the growth and isolation of heterotrophs (due to the inclusion of yeast extract) and is thus not elective for methylotrophs. Also for an industrial process an organism that does not require growth factors (such as vitamins) is preferable. Medium C might well enrich for methylotrophs requring such expensive growth factors. In practice, pdms such as described in 3) would be carried out over different periods of time, and at varying pH and temperature. Environments most expected to contain methylotrophs are swampy areas, where methane would be produced and then be oxidised to methanol. The organism finally chosen for the process we are considering here was isolated from a waterlogged soccer pitch! 4.9.4 Stage Two - Lab Scale Once methanol-using organisms had been isolated they were screened in small-volume shake-flask cultures to determine their ability to grow in methanol-minimal-medium (such as Medium B described in the previous section) to produce high yields at high growth rates. Optimum growth temperatures and pHs were also determined. Those organisms growing with high rates and yields and at relatively high temperatures (> 30°C) were selected for further study (growth at elevated temperatures would reduce cooling requirements). Studies on selected organisms were carried out in 5 litre continuous cultures to determine biomass yield coefficients, maximum growth rates (CLmU), affinities (Ks) for methanol and oxygen, and stability (of morphology and culture characteristics) on prolonged culture. Suitable cultures were identified. Shake flask cultures n Answer true or false to each of the following statements. For SB production continuous rather than batch cultures are preferred because: batch and continuous CUlturDS 1) 2) 3) continuous culture systems give higher outputs for a bioreactor of given size; continuous culture systems are easier to operate; continuous culture systems give more control over the culture environment and hence biomass quality; continuous cultures are less susceptible to mutation; continuous cultures convert higher proportions of substrate. 4) 5) Our response: 1) True. Batch cultures give lower overall outputs than continuous cultures, as they suffer from non-productive down-time (the time taken to empty, clean, re-sterilise and re-fill the fermentor). After inoculation, considerable time can be taken for biomass to build up to a level where substrates are effectively utilised. Continuous cultures do not suffer such drawbacks once they are in operation. False. Continuous culture systems are more difficult to operate than batch cultures. Medium is continuously added and withdrawn, making the pmss more prone to contamination. Maintenance cannot be carried out during len@hy culture runs, making the equipment more prone to breakdown. Batch cultures do not suffer such drawbacks. 2) 92 Chapter 4 improved economics animal testing obligate meth ylotroph design flexibility True. In batch cultures, at the end of the growth phase when biomass concentration is high, the environment changes rapidly (as nutrients are becoming rapidly depleted) and growth rate decreases. Environment and growth rate can affect properties such as the protein content of cells, and thus biomass quality can vary from batch to batch. In continuous cultures the environment is constant. The specific growth rate (p) is fixed by the dilution rate (D) so the product is less subject to variation. In fact at steady state p = D. False. Continuous cultures operate for lengthy periods. Spontaneous mutations will arise and if they can compete successfully with the parent organism (by virtue of higher growth rate) they can predominate in the culture. Batch cultures have short growth times and so do not suffer such drawbacks. False. Batch cultures can convert high proportions of substrates, as growth can be allowed to proceed until substrates are exhausted. In continuous cultures substrates are never fully converted, as medium is continuously removed. In fact, residual substrate concentration increases as the dilution rate increases, until virtually all of the medium remains unused. Continuous cultures usually recycle the medium after biomass removal to inmase the efficiency of substrate conversion. Item 1 is the major factor in choosing continuous systems rather than batch systems for SCP production. Economics are improved by lower capital cost for the bioreactor (a major equipment cost, see Section 4.10) and by a higher output rate. Item 3 leads to greater control of product quality. Cultures with high affinities of G and methanol were checked for amino acid profiles, protein and nucleic acid content, and lack of pathogenicity and toxicity in rats. Details of these toxicity tests are given in the resource material at the end of this chapter. 4.9.5 Stage three - 1 m3 scale Methylophilus methylotmphs was the organism selected for further study. The organism is an obligate methylotroph, which means that it can use only C1 organic molecules as a source of carbon. Suitable strains of organism were cultured in aseptic pilot-scale 1 m3 continuous cultures to ensure stability at that scale in long-term cultures, and to produce material for extensive feeding and toxicity trials. It was important at this stage to ensure that there could be future flexibility in design and operation of a commercial-scale plant, so chemical composition, nutritional value and toxicological effects of the SCP product were examined with respect to variations in plant design and operation. What aspects of plant design and operation would you have chosen to study at this stage in order to give maximum flexibility in design and operation of a future commercial-scale plant? The factors chosen were variations in vessel size, the effect of non-aseptic operation and medium recycle, and variations in recovery and drylng procedures. In parallel with these studies, developments were underway to find the most economical method of large-scale culture and down-stream processing. The biochemistry of methanol and ammonia utilisation by MethyZqhi1u.s methylotrophus was also studied to pinpoint possibilities for manipulation. n n Why would you expect Methybphilus methylotmphus to be non-pathogenic? Single cell protein 93 mass transfer heat transfer mixing As an obligate methylotroph, the organism would not be able to grow in animals, where substrate would not be available. Preliminary trials showed that Methylophihs methylotroph was a suitable source of feed. Changes in culture system, operation, mode of recovery and drying did not adversely affect this. The development team was confident that the pilot-scale and production-scale facilities could be developed effectively. 4.9.6 Stage four - 40 m3 scale It was expected that when scaling up the fermentation to about 40 m3 the parameters most likely to be affected would be mass (oxygen) transfer, heat transfer and mixing. On the basis of the growth reaction, which can be written as CH30H + NH3 + 02 > biomass + COZ + H20 and assuming conversion of methanol to either biomass or COz, the relationship between (i) yield and 02 requirement, (ii) yield and minimum oxygen transfer rates (OTR), (iii) yield and heat production, and (iv) productivity and heat production (with various yields) can be estimated. You will find these data in the Resource Material at the end of this chapter. The ideal requirements for the 40 m3 system and projected larger production system are listed below. Study them all, then give reasons why you think each one was considered necessary. Continuous operation. Aseptic fermentation. Optimum output. An alternative to centrifugation for harvesting biomass. Medium (water) recycle. Aseptic harvesting procedure. Appropriate drying procedure. Our responses to these are: 1) If you cannot answer this question refer back to Section 4.9.4. 2) Medium recycle (item 5) means that organic materials released from MethyloPhilus can build up in the medium. These are available for the growth of contaminants. The process cannot be run at low pH to suppress the growth of contaminants as the organism itself requires a pH of about 7.0. Aseptic operating conditions are therefore required. Continuous cultures can be run at different dilution rates. An optimum rate must be chosen, taking account of output, substrate utilisation and aeration and cooling requirement. We will work this out for ourselves later on. 3) 94 Chapter 4 Centrifigation is too expensive a method for the bulk recovery of bacterial cells, due to their low density. Medium recycling is requved to save water, save unused nutrients, and save on waste treatment costs (spent medium contains unused organic nutrients and so would otherwise have to be treated to reduce BOD). Aseptic harvesting is necessary to overcome the need for rnedium R-sterilisation before recycling. Sterilisation costs are high. If biomass can be recovered by an aseptic process, the medium can be recycled without re-sterilisation. This excludes centrifugation, which cannot be operated under aseptic conditions. Drylng procedures need to be effective and economic. 4) 5) 6) 7) The bioreactor We will now deal with quite sophisticated issues concerning reactor design and performance. Although we deal with these in a sympathetic manner, you may find it useful to refresh your knowledge by reading more about bioreactors. The BIOTOL series offers opportunities to learn more about fermentors. With the aid of the data given in the Resource Material at the end of this chapter, answer the following. (Refer especially to Figures 4.10,4.11,4.12 and 4.13). A 36 m3 culture is grown at D = 0.3 h", at a biomass concentration of 30 kg m-3 The biomass yield coefficient is 05 kg dry biomass per kg methanol. Estimate: 1) 2) 3) 4) the output (kg biomass h-'1; the minimum OTR (kg 02 requwd m" h-'); the heat evolution rate (k Joules m3 h-I); the concentration of methanol requwd in the incoming medium to support a biomass concentration of 30 kg m-3 - assuming 90% utilisation at D = 0.3 h-'. air-hft PmS~mcY~ reador For a 40 m3 pilotscale fermentor an air-lift pressure-cycle bioreactor was chosen. The reactor (Fiy 4.8) has a working volume of 36 m3 with a 42 m high riser, fed with up to 80,OOO m air h ' at pressures of about 3 x lo' Nm-' (about 3 atmospheres). A maximum OTR of 10 kg 02 m-3 h-' is produced. The riser section is baffled at intervals to slow the circulation rate (and so enhance the OTR by increasing bubble residence time) and to promote bubble break up (again enhancing OTR). The rise of expanding bubbles is the sole source of energy input to the bioreactor, producing lower densities within the riser and thus causing liquid circulation down two downcomers. Bubbles coalesce in the top of the riser, their slip velocity (the rate at which they rise through the medium) thus increases and they disengage (break free of the liquid). slip velocity Single cell protein Figure 4.8 Pressurecycle reactor. n Answer true or false to each of the following statements. Compared to a conventional stirred bioreactor, an air-lift pressurecycle bioreactor is more suitable for development of a SCP process from methanol because: 1) 2) 3) 4) it is easier to construct and maintain at all scales; the system gives higher OTRs and at greater energy efficiency; it is easier to operate on a large scale because of defined and controllable flows; it gives more even distribution of nutrients so that the culture is more likely to obey chemostat kinetics; pilot and production-scale bioreactors of this type behave more like conventional stirred lab-scale bioreactors than equivalent-sized conventional bioreactors, thus making scale-up from lab-scale more reliable. 5) Our responses are:- 1) True. Stirred bioreactors require finely engineered aseptic seals on shaft bearings, which require maintenance. The larger the bioreactor the greater the problem- Air-lift bioreactors have no such moving parts. 96 Chapter 4 True. Pressurecycle systems need higher power input for air compression, but this is more than compensated for by saving on power for stirring. Air filtration costs are reduced because air is delivered at higher pressure, so actual volumes filtered are lower. Oxygen transfer is promoted by the high hydrostatic pressure generated at the base of the fermentor (at which point the air is introduced); Such systems give OTRs of up to 10 kg Ozm” h -’ for power inputs of 1-2 kW m- . True. Pressurecycle bioreactors have controllable and predictable flow patterns, which makes scale-up more predictable. Factors such as OTR and heat transfer are easier to arrange at large scales. 2) 3) 4) False. Pressure cycle bioreactors do not give even distribution of nutrients. If air and nutrients are introduced at a single point then 02, CO2 and nutrient concentration, as well as hydrostatic pressure, change in a cyclic manner as the medium flows around the reador. 5) False. The behaviour of stirred bioreactors does not resemble closely pressure-cycle fermentors at any scale. The characteristics of this pilot-scale system were extensively studied to discover how changes in medium viscosity, surface tension, flow rate and sparger and baffle design affected bubble size, slip velocity, disengagement and OTR. Bubble size and slip velocity proved to be crucial, as bubbles needed to be 1-10 nun in diameter for optimum gas exchange. Large bubbles (10 nun or more) produce inefficient gas exchange (with low surface:volume ratios and short residence times due to high slip velocities). Smaller bubbles (less than 1 mm diameter) fail to coalesce and disengage at the top of the riser. Baffle design was adjusted to give appropriate bubble sizes with thorough lateral mixing, but avoiding high shear and high turbulence (which cause formation of bubbles of less than 1 mm). High rates of gas disengagement are achieved in an enlarged unbaffled section of the fermentor at the top of the riser. n riser promotes bubble disengagement from the medium? In the riser, baffles are placed at intervals to break up bubbles by increasing turbulence and shear. At the top of the riser the expanded section decreases the upward flow rate of the medium and this, together with the lack of baffles, decreases turbulence and shear, which in turn promotes coalescence of bubbles. Larger bubbles form which have increased slip velocity, so they more easily disengage from the medium. Bubble behaviour was studied in the pilot-scale bioreactor so that a complete model of flow, OTR, mixing, cooling, energy requirement and disengagement could be developed for this system and larger production-scale vessels of similar type. The yield problem As well as understanding and optimising the working efficiency of the bioreactor it was also vital to optimise the efficiency of the fermentor - the micm-organism itself. The biochemistry of methanol utilisation is similar to that of methantrophs described in Section 4.9.2, and is shown in Figure 4.9. controlling bubMesize baffles How do you think that the enlarged unbaffled section at the top of the fermentor Single cell protein 97 Figure 4.9 The utilisation of methanol as a carbon and energy substrate by methylotrophs. Despite optimised culture conditions, yield coefficients for methanol were lower than the expected level of about 0.5. Yields were, in fact, lower than those achieved routinely with the lm3 fermentation system The problem was traced to the cyclical nature of the pressurecycle system, which, due to introduction of air, ammonia, methanol and other nutrients at single points, leads to cyclical changes in nutrient concentrations as well as changes in pH, temperature and hydrostatic pressure. These parameters change in cycles as the medium flows around the reactor. Such changes do not occur in stirred vessels, accounting for differences in yield observed in comparison with those in the '1 m3 system. Experiments showed that the cyclical changes in methanol concentration were the cause of reduced yields. It was assumed that this was due to an hase in maintenance energy. ~aintenance energy is the energy needed to maintain cells. It is energy that is used for cell repair, maintaining osmatic balance, transport etc. It is not used for growth. n in methanol concentration in the pressurecycle bioreactor? dud yields maintenance ene~ How would you overcome the problem of low yield caused by cyclical changes 98 Chapter 4 The solution was to introduce methanol into the medium at many points through nozzles in the bioreactor wall. In this way even methanol distribution was achieved, and yields of 0.5 kg biomass per kg methanol can be obtained. Genetic engineering was also carried out to improve yields even further. The production organism, strain AS1, assimilates ammonia according to the scheme: genetic engineering glutamate sYn*etase Ammonia is accepted by glutamic acid in an energy (ATP consuming) step, and converted to glutamine. This reaction is catalysed by glutamate synthetase (GS) and can be written as: GS glutamic acid + NH3 + ATP > glutamine + ADP + Pi. glutamate OXodutamate amino- Glutamine then reacts with a-oxoglutarate to form 2 molecules of glutamic acid. This reaction is catalysed by glutamate oxoglutamate aminotransferase (GOGAT) and can transferase be written as: glutamine + a-oxoglutarate + NAQP)H GoGAT> 2 glutamic acid +NADP)' One molecule of glutamic acid is recycled to accept ammonia and the other becomes available for protein synthesis. Other bacteria such as E. coli assimilate ammonia by incorporating it directly into a-oxoglutarate in a reaction catalysed by glutamate dehydrogenase (GDH). This reaction can be written as glutamate dehydrogenase GDH a-oxoglutarate + NAD(P)H + NH3 > glutamate + NADP)+ This b -passes the ene consuming GS step of the GS/GOGAT pathway used by It was decided to use genetic engineering to put the more-efficient GDH system from E. coZi into Methylophilus methybirophus AS1. The strategy was as follows: Methy 7 ophilus methylotrop % and is thus more efficient. Single cell protein 99 select GOGAT-deficient mutants of AS1 after treatment with nitrosopaniche; (these mutants would not be able to assimilate ammonia). isolate the GDH gene from E. coli. insert the GDH gene into a plasmid vector using the sal I restriction enzyme; 0 transfer the plasmid to the GOGAT deficient m&mt of AS1; 0 ensure the GDH gene was expressed in the new host. The plasmid vector was chosen for its ability to be transfed into bacteria of different species. It was successfully transfed to the GOGAT-deficient mutant of ASl. These organisms then became able to assimilate ammonia and grow at normal growth rates. Yields of the engineered organism were about 5% greater than the parent organism in the pilot-scale culture system. strategy n Select the correct ending to the following statement. It is necessary to introduce the GDH recombinant plasmid into a GOGAT-deficient mutant strain of AS1 because: 1) GOGAT would otherwise compete with GDH for ammonia; 2) 3) GOGAT would otherwise compete with GDH for a-oxoglutarate; GOGAT would otherwise compete with GDH for glutamine. The correct answer is 2). Both GDH and GOGAT use a-oxoglutarate as a substrate. GOGAT deletion ensures a-oxoglutarate is directed only through the GDH step. Although the genetically engineered organism produced higher yields on methanol than the parent organism, it was not used for routine production. This was mainly because the legislation governing the large-scale production of genetically engineered organisms and their use as feed was, at the time, under review and was uncertain. It was therefore easier to establish a process based on a 'natural' organism. Sterile Engineering Sterility was essential in the bioreactor and in the biomass separation area, so that liquid could be recycled without resterilisation. This was achieved by careful design and operation. Cell recovery by centrifugation would have been expensive and difficult to achieve without risk of contamination, so a 2-stage process was developed. Stage 1, run under aseptic conditions, involved heat lysis of cells at 70°C followed by acidification. This caused flocculation of cells around gas bubbles released from the liquid, forming a froth of cell material which was skimmed off. This achieved a 10-fold thickening to 10-12% w/w. The froth is taken off through a seal to a non-aseptic, but clean area, to Stage 2, where it is centrifuged to a 20% w/w suspension for drying. An economic drying process was developed at this stage. Mislation heat lysis and acidification 100 Chapter 4 pnrteen By this lime, the product had proved, in continuing toxicity and feeding trials, to be safe and effective feed material.The p"xess was patented and the tradenamePruteen given to the product. A licence to sell the product as feed was granted by the Ministry of Agnculture, Fisheries and Food. 4.9.7 Stage Five - The Pruteen Production Process The production-scale fermentation unit, with a projected annual capacity of over 50,oOO tonnes was fully commissioned in 1980. The bioreactor (Figure 4.8) is 60 m high, with a 7 m base diameter and working volume 1,500 m3. There are two downcomers and cooling bundles at the base. Initial sterilisation is with saturated steam at 140°C followed by displacement with heat sterilised water. Air and ammonia are filter sterilised as a mixture, methanol filter sterilised and other nutrients heat sterilised. Methanol is added through many nozzles, placed two per square metre. For start-up, 20 litres of inoculum is used and the system is operated as a batch culture for about 30 h. After this time the system is operated as a chemostat continuous culture, with methanol limitation, at 37°C and pH 6.7. Run lengths are normally 100 days, with contamination the usual cause of failure. produdon boreactor Harvesting is continuous and under aseptic conditions, with medium recycle without re-sterilisation. Residual nutrients are monitored to ensure there is no build up (for example of phosphate). The cream (biomass) from flocculation and centrifugation is adjusted to normal pH (remember it was acidified for flocculation), then dried in an &-lift flash drier. For this the mam is fed into the venturi of the drier tube and is carried upwards by the hot drymg gas, after which it is recovered in high efficiency cyclones. The product is cooled and the very fine material (waste dust) is separated and remixed with cream from harvest for redrying. The granular product is lightly coated with vegetable oil to suppress dust and is stored pending quality control analysis. The product is sold in the granular form or is ground to a fine powder. downstream P-Saing Draw a flow diagram of the Pruteen process. I SAQ4.10 I Single cell protein 101 The organism grows on methanol with a yield coefficient of 0.5. The maximum growth rate (p,,.,& is 0.55 h”. Theoretical considerations show that, with a methanol concentration of 70 kg m3 in the incoming medium to the bioreactor, the biomass concentration and residual methanol concentration (in outgoing medium) would vary with dilution rate as outlined below. Dllution rate (D) (h-’) Biomass concentratlon ResMual methanol (kg dry biomass m”) 0.1 34.3 1.4 0.2 33.1 3.7 0.3 31.1 7.8 0.4 26.3 17.3 0.5 2.5 65.0 concentratlon (kg m”) For each dilution rate calculate: 1) the productivity (kg biomass produced m-3 h-’); 2) the percentage methanol utilisation; 3) the minimum OTR (kg oxygen required m” h-’1. The bioreador has a maximum OTR of about 11 kg 02 m” h”. From these data select the optimum working dilution rate for the bioreactor. For a 1500 m3 system, working at D = 0.2 h-’ and biomass concentration 36 kg m-3, with a yield coefficient of 0.5, calculate the annual: 1) output; 2) methanol requirement. Listed below are some of the steps in the development of an SCP process. List these steps in the sequence in which you think they would be performed, starting with the first and ending with the last. If you think that any steps would be performed concurrently (at the same time), list them side by side. Marketing. Measure protein content and amino acid profile. Full-scale production. Isolate organisms capable of using the substrate. Apply for sales licence. Measure affinity for substrate. Perform feeding trials in animals. Establish the size of the market. Measure temperature optimum (for organisms’ growth). 102 Chapter 4 4.9.8 The Outcome The 1973 oil price rise caused an increase in the price of fuels, including the natural gas substrate for the process. However, this was offset by corresponding price increases in competing feedstuffs, mainly soya and fish meal. The high value of soya then stimulated increased production and in 1979 soya was over produced, leading to a collapse in the price of protein feeds. In 1983 the price of soya in Europe was below its 1973 level, whereas oil prices had again risen sharply in 1980. Pruteen was then selling at $600 per ton, more than twice the price of the conventional soya feed. In an attempt to maintain the economic viability of the process the product was promoted as a milk substitute (milk powder for animal weaning being more expensive than solid feeds like soya). However, the size of the market for milk substitutes was relatively small, and for a time the plant was operated at reduced capacity. By 1986 the price of methanol was $179 per ton, which at yield factors of 0.5, contributed $358 per ton to the cost of the product. This in itself exceeded the price of soya for feed, and the production plant was closed down. The cost of development of the Pruteen process has been estimated at $180 million (1986 prices). Despite the lack of economic success, the process has not been judged a failure. Extended feeding trials throughout the development and production stages showed that Pruteen was a satisfactory form of protein feedstuff, compared to conventional sources, and was without toxic side effects. The technology that had been developed for the Pruteen process has been sold under licence to other companies in the UK and elsewhere for SCP production (for example in Myco-protein production). A version of the pressure cycle bioreactor has also been developed for effluent treatment. ICI have also applied the technology to other biotechnological products, such as biological feedstocks for plastic production. 4.10 Economics of SCP production Detailed economics of individual industrial processes, including SCP processes, are usually regarded as confidential, out of fear that publication may lend advantage to competitors. In addition, 'economy of scale' rule generally applies (that is as the production capacity increases, the cost of the product decreases), so that direct comparisons can only be made between systems of similar capaaty. Some economic data on SCP processes have been published and are presented in the Resource Material at the end of this chapter. You should appreciate that the data are outdated by more than a decade, during which time substrate costs will have varied dative to each other, and technology will have improved. This means that the comparative costs presented in Table 4.13, for example, may not be now as presented there. Nevertheless the data presented do provide an outline of the economics of SCP production. The processes referred to in the Resource Material are not necessarily those mentioned in the text and so you may find some differences in detail. changing market mnditions success from the technological development confidential data Single cell protein 103 Use the data given in the Resource Material to complete the following summary of economics of SCP production. Delete inappropriate choices and replace Xs with figures. (You will find Tables 4.9 - 4.15 most useful in making your choices). 1) The most significant production cost in SCP production is the cost of raw materials/labour/running costs, ranging from X to X% of total production cost. The substrate contributes least to production costs when it is a waste/non-waste. For liquid substrates the most significant equipment cost is medium preparation/fermentation/harvesting/drying, ranging from X to X% of the total running costs. The most significant running cost (for liquid processes) is inoculum preparation /medium preparation/fennentation/ harvesting/drying, ranging from X to X% of the total running costs. The most sigxuficant running cost of fermentation is sterilisation/ aeration/cooling, ranging from X to X%- The most significant cost of fermentation equipment is the air compression system/the bioreactor/ the sterilising system/the cooling system. The most significant cost of harvesting and drying equipment is drying/ filtration/centrifugation. The most sigruficant running cost of harvesting and drying is drying/filtration/centrifugation. 2) 3) 4) 5) 6) 7) 8) Use the data in the Resource Material to answer the following question. It is 1977. The bacterial SCP from methanol plant referred to in Table 4.9 does not produce protein at a price that competes with soya protein. By how much would the cost of methanol have to fall in order that the protein from such a plant can be produced competitively with soya protein? You can assume i) that the SCP processes referred to in Tables 4.7 and 4.9 to 4.15 are of 2 x Id tons annual capacity, ii) that yield on methanol is OSkg biomass per kg methanol, iii) bacterial SCP contains 60% protein. 104 Chapter 4 For the yeast process from n-alkanes, which of the following contributes most to the cost of yeast production, and which contributes least? (Tables 4.7; 4.11 and 4.12 are helpful). 1) Aeration (during fermentation). 2) Cooling (during fermentation). 3) Ammonia. 4) Labour. 5) Drying. You are developing a process to produce SCP from a carbohydrate (sugar) substrate. The price of this substrate is $0.2 per kg sugar. You have a yeast, Cmzdida ifulika, which is suitable for food or feed and which grows on the sugar substrate with a yield factor 0.48. You have also a filamentous fungus, Fusarium imh, which is also suitable for food or feed and which grows on the sugar substrate with a yield factor of 0.45. Both fungi will grow at pH 2.5, at which non-aseptic processes can be operated (that is without sterilisation). However, the SCP grown in non-aseptic systems is suitable only as feed. The SCP from both organisms can be used as a high-protein food additive, but Fusmium sp. must be ground up (powdered) for this. In addition, the filamentous fungus can be used to make meat substitutes. For this the SCP must be prepared deep-frozen and not dried. The selling price of the SCP must be the same as or less than competing food and feedstuffs The price of conventional competing protein feeds is $0.80 per kg protein. The price of conventional competing high-protein food additives is $1.55 per kg protein. For a meat substitute, the SCP can be priced at $1.05 per kg biomass. An existing fermentation unit has been made available for use (it would be the same for both organisms). Thus there are the following fixed costs for production of either organism. Contrlbutlon to produdlon cost (S per kg biomass) Substrates other than the 0.1 carbohydrate substrate Fixed running costs 0.055 (including fermentation) Labour 0.05 Total: 0.205 Single cell protein 105 Thus, for the production of either organism you have, apart from the cost of the carbohydrate substrate costs of $0.205 per kg biomass. Assume both organisms give the same output. You have the choice of whether or not to operate an aseptic or non-aseptic system, and which down-stream processing operations to use. The costs of these operations are given below. Contribution to production cost ($ per kg blow) Sterilisation 0.04 Centrifugation (with washing) 0.055 Dewatering (filtration with washing) 0.001 Drying 0.02 Deepf reezing 0.04 Milling (grinding) 0.01 Which of the following products could be most profitable? 1) cmtdida sp- as feed. 2) 3) Fusarium sp. as feed. 4) 5) Cmtdida sp. as high-protein food additive. Fusarium sp. as high-protein food additive. Fusarium sp. as meat substitute. 106 Chapter 4 Summary and objectives A variety of SCP processes have been developed with a view to producing food and feed from alternative or waste carbon sources. Processes have been based on CG, carbohydrates, hydrocarbons and their derivatives. Technology has ranged from relatively simple open lagoons or solid-substrate fermentations to largescale aseptic continuous cultures in fernentors. Feed processes have been developed mainly in Europe, Japan and the former USSR, where feedstuffs are in short supply. Processes have not generally been successful due to unfavourable economics (rising substrate costs and decreasing cost of soya), or political and social pressures (particularly against SCP produced on oil-based substrates). In the former USSR, where different economic and political systems applied, feed processes were operated on a large scale. For food, SCP has been produced on a smaller scale and is mainly limited to yeasts, the use of which in food has been traditional. The exception is Myco-protein, which is being promoted as a health food, rich in protein and lacking animal fat, and which can be used as meat substitutes in high-value vegetarian convenience foods. Industrialised countries that have developed the technologies are not the countries suffering population explosions and food shortages. The so-called developing countries that have the population and food problems, do not generally have the industrial base and technological expertise necessary to operate large-scale processes. These factors limit the application of SCP processes. Perhaps SCP will become a thing of the future, when conventional proteins might be in short supply in industrialised countries. Unf~rt~~tely we might not have to wait long for this situation to arise. 1988 and 1989 saw drought conditions over significant parts of the USA - a reminder that climate is not fixed, but undergoes changes. Fish stocks in many oceans are not managed or conserved. At least we can draw comfort from the fact that SCP technology has been tried and proven, and is available if need be. After studying the material in this chapter, you should be able to: describe the advantages of using micro-organisms as food and feed, compared to conventional protein sources; describe the different physiological types of miao-organism used in SCP production; use knowledge of the organisms and technologies involved in the development of SCP processes to make valued judgements on alternative strategies for process development; compare the economics of SCP processes with those of conventional protein production. Single cell protein 1 07 Resource Material Toxicological tests on SCP in animals To test for possible toxic effects, rats, mice, dogs, chickens pigs, and fish were fed Pruteen at various levels in their diet, ranging from 30% to 60%. Control animals were fed casein (milk protein) instead of Pruteen. Feeding periods ran@ from 28 days to 3 years. The parameters measured are outlined below. Feed intake rate. Growth rate. Haematolcgical characteristics (haemoglobin level, red blood and white blood cell counts, platelet count, blood clotting time). Biochemical characteristics (plasma levels of alanine and aspartate transminases, alkaline phosphatase, triglycerides, cholesterol, urea, uric acid, allantoin, glucose, protein, albumin, sodium, potassium, calcium, magnesium, phosphorus; urine levels of protein and glucose). Histopathological examination (up to 40 different tissues were post-mortem and examined for abnormality). Effects on reproduction [size of litters/broods; evidence of teratogenicity (physical defects) in foetuses]. 108 Chapter 4 Oxygen requirements and heat outputs of organisms grown on methanol Figure 4.10 Typical effects of yield coefficient on oxygen requirement when only biomass and C02 are produced (methanol as substrate) Figure 4.1 1 Typical relationship between minimum oxygen-transfer rates and yield coefficients at various productiiities (methanol as substrate). Single cell protein 109 Figure 4.12 Estimated relationships between heat production, calculated from heats of combustion and yield coefficients (0 )or calculated from experimentation (0). Methanol as substrate. Figure 4.13 Estimated relationships between heat production and productivity at various yield coefficients 110 Chapter 4 Economics of SCP Production $ kg-1 % Protein $ per kg proteln Beef 1.54 15 20.3 Pork 1.10 12 19.1 Pouttry 0.66 20 6.6 Cheese 0.78 24 6.5 Milk Powder 0.46 36 2.5 Soya Flour 0.1 5 52 0.6 Peanut Flour 0.1 5 59 0.5 Yeast (from 0.42 53 1.4 n-alkanes) Yeast (from 0.33 53 1.3 molasses) Fungi (from 0.15 43 0.7 cellulosis) Algae 0.66 46 2.8 Table 4.7 The production cost of various proteins (1 980 figures, USA, prices in $ US). 10' tonnes per year capacity, 1975, Japan, producing yeast investment cost 1 1.2 0.85 1.1 Production cost 1 1.4 1.6 1.3 10'tonnes per year capactty, 1975, Israel, producing yeast Investment cost 1 0.68 0.77 Production cost 1 0.75 1.24 nAlkanes Methanol Ethanol Molasses n-alkanes Methanol Molasses 2 x lo6 tonnes per year capacity, 1977, USA Yeasts, Bacteria, mAlkanes Methanol Investment cost 1 Production cost 1 0.73 0.98 I Table 4.8 Cost of SCP grown on various substrates in comparison to the n-alkane process. Single cell protein 111 The tables that follow give the costs of various SB production processes in comparative rather than in actual form. To see what this means examine Table 4.9. The production cost of raw materials for yeasts grown on n-alkanes is given as 585. This means that the cost of raw materials accounts for 585% of the total production costs of this process. The same cast for bacteria grown on methanol is 73.8. This means that in this case 73.8% of the total production cost is accounted for by raw materials. This does not mean that the actual cost of raw materials for the methanol process is more than that for the n-alkanes process, as the total costs of the two processes are not necessarily similar. y-1 Bacteria, Yeast, Ethanol Fungus, malkanes Methanol Sulphlte waste Ilquor Depreciation (of value of production plant) Raw matrials (total) a) Carbon substrate b) Ammonia c) Phosphoric acid d) Mineral salts e) Other Labour 9.3 58.5 29.4 11.1 9.9 2.9 5.2 8.4 5.8 5.8 73.8 77.1 47.4 63.9 11.8 3.2 12 4.8 2.6 1.9 3.3 6.2 5.1 9.1 55.1 17.0 16.2 13.3 4.2 4.4 11.0 I Table 4.9 Relative equipment costs of various SCP processes. I Yeast, malkanes Bacteria, Yeast, Ethanol Fungus, SulphL Methanol waste liquor ~~ Supplies Storage 3.3 4.9 4.0 4.9 Medium 1.1 1.5 1.2 1.6 Preparation Inoculum 0.9 1.1 0.9 1.3 Preparation Fermentation 51.2 43.4 50.0 50.6 Harvesting 14.2 11.0 13.1 7.6 Drying 17.8 23.1 18.7 17.0 Product storage 11.5 15.0 12.1 17.0 Table 4.1 0 Relative production costs of various SCP processes. 112 Chapter 4 I y-9 Bacterla, y-, Fungus, nAlkanes Methanol Ethanol Sulphite waste llquor Sterilisation 21.4 Aeration 88.9 70.9 86.3 92.3 Cooling 7.9 5.1 12.5 6.7 y-, Bacterla, Yeast, Fungus, Sulphitc mAlkanes Methanol Ethanol waste llquor Supplies Storage 1.2 0.2 0.2 0.2 Medium 1.6 2.4 1 .o 1 .o Preparation Inoculum 1.8 2.5 2.3 2.6 Preparation Fermentation 67.9 61.7 53.5 76.8 Harvesting 5.9 3.8 16.0 3.6 Drying 21 .o 28.8 26.4 15.1 Product storage 0.6 0.6 0.6 0.7 Table 4.1 1 Relative running costs of various SCP processes (The cost of the electricity, fuels and water used in various processes). Table 4.12 Relative running costs of fermentation of various SCP processes. y-, Bacterla, Yeast, Fungus, Sulphitc mAlkanea Methanol Ethanol waste llquor fermentor 40.1 34.5 42.1 34.4 Air supply system 20.0 21.2 16.4 24.4 Cooling system 37.4 32.0 40.6 39.1 Sterilisation 10.6 system Others 2.5 1.7 0.9 2.1 Table 4.13 Relative costs of fermentation equipment for various SCP processes. Single cell protein 113 Yesst, mAllcanes Bacteria, Methanol Yeast, Ethanol Flitration 11.9 Centrifugation 26.8 31.1 27.7 Drying 55.6 68.9 57.4 Others 5.7 14.9 Table 4.1 4 Relative costs of harvesting and drying equipment I Yeast, mAIkenes Bacteria, Methanol Yeast, Ethanol I I I Filtration 7.4 Centrifugation 10.0 11.7 Drying 77.7 88.3 Others 4.9 8.5 61.8 29.7 Table 4.15 Relative running costs of harvesting and drying.