Comparison of alternative design approaches

Contents - Previous - Next

Table 7.7 presents digester sizes calculated according to criteria for Indian (KVIC), USEPA (1979), Mullan et al. (1984), and this report. The 27°C temperature is based upon current Indian practice. Both the Indian and USEPA approaches result in larger and, from an operating stability standpoint, conservative sizes.

Table 7.7: Comparison of Calculated Digester Sizes Operating at 27°C Ambient Temperatures.

a) Based on ESCAP (1980) using 27°C;
b) Based on Hashimoto equations using 27°C;
c) Based on equivalent size assuming 80 g per capita per day solids from 90 people weighing 4,500 kg;
d) Assumes 4,500 kg total cattle weight (average 450 kg per animal;
e) Assumes 4,500 kg total live weight, manure collected fresh and flowing by gravity to a digester (no water added);
f) Mullan et al. 1984.

Problems and solutions of feedstocks and effluents in full-scale biogas plants (based on hobson, 1987)

Digester feedstocks
Mechanical problems
Chemical and biochemical treatments of feedstocks
Digester effluent

One of the main advantages of applied anaerobic digestion is that results from small scale experiments can be reproduced in a full-size plant. Thus, many aspects of the overall design (including the type: stirred-tank, UASB, etc.) and running of a digester system may be based on laboratory and pilot-plant experiments and scaled up. Some feedstocks used as charging materials to the digestion system can be digested without problems while others can not. The solution of the problem of the resistance to digestion of a specific material may lie in further small-scale experiments, or it may mean modification and experimentation with the large-scale plant. The following section deals with the problems of feedstocks that are difficult to digest, and some solutions to this problem.

Digester feedstocks

The physical and chemical state of a material feedstock used for anaerobic digestion is initially determined by its source. The feedstock may be a clear liquid, a suspension of solids in a liquid, or a "solid" - a material with less than 70 - 80% water content. Various digester systems have been designed to treat these physically different forms of feedstock. Sometimes physical and chemical problems may be solved by modifying the feed. These modifications may allow the original type of digester to be used, or they may be such that a different type is needed.

Mechanical problems

Slurries Biodegrading of a feedstock material, to determine the feasibility of digestion of a particular waste, and the basic parameter" of biogas production, can usually be carried out on laboratory or small pilot plant scale. There may almost no problems in the use of wastewaters with dissolved substrates (except some very oily materials). However, mechanical problems in pumping and piping liquid wastes containing large suspended solids may dictate the minimum size of digester which can treat the raw waste. Obviously, chopping, maceration, or separation of large particles, to produce a suspension in liquid which can pass through small pipes, can profoundly alter the kinetics of solids digestion. Therefore, small-scale results with macerated solids may not be representative of large scale results with the raw feedstock. In small scale experiments, batches of feedstock for the digesters are usually collected from the main source. It is again obvious that these batches should be representative of the proposed large scale feedstock. Biodegradability experiments (Goering and Van Soest 1970; Jewell 1976; Kimchi 1984) were found to be a good tool for estimation of the extent a specific feedstock will be degraded.

Low solids wastewaters can be piped to digesters, as in any chemical process plant; high-solids sludges and slurries are different. Problems have become particularly apparent with cow- farm manures. Farm slurry systems are generally designed to move large volumes quickly. Most farm digesters are relatively small, and have long retention times. This means that small volumes of slurry have to be pumped and piped at low overall rates. The farm slurries contain grit, animal hairs and fibrous feed residues (straw) and these are incompatible with the use of the small bore pipes and pumps appropriate to the low slurry volumes fed daily to the digester. Pipes need to be 7.5 cm or more in diameter, as straight as possible, with any bends of large radius, and with full-bore valves, and must have provision for access to clean the line with a stick. Mono-type pumps are satisfactory (but rotors are amortized quickly), and their ability to run in reverse aids clearing of blockages (Hobson and Feilden 1982; Summers et al. 1984). They are suitable for feedstocks without stones or straw. The low daily flows required are obtained by intermittent pump operation for short periods. Similar pipe and pump systems are needed for the effluent of pumped outflow digesters, although much of the larger particulate material is disintegrated during digestion. Grit from animal house floors may not cause much problem with large-bore pipes, but it does increase the rate of wear of pumps, particularly centrifugal types. Larger stone" can cause blockage and breakage problems, as was found with some digesters taking feed from cattle feedlots with earth-floor enclosures (a good example is the large scale experiment in Oklahoma). Settlement, or coarse screening of the slurry may be needed in such cases. Settling of grit and other undigested materials occurs in the bottom of the digester, and reduces the available volume for digestion, thus reducing the HRT.

The published information on the engineering properties of farm wastes and slurry pumping and flow through pipes (e.g. Howard 1978; Chen 1982; Bohnhoff and Converse 1987) is relatively sparse, since companies keep such information as proprietary. Much of the building of farm digester piping has been done by trial and error, and the information is kept with the designers. This is to some extent unavoidable, as slurries that are nominally the same vary with animal feed and farm conditions, and there are large variations between pig, cattle and poultry waste slurries. Much also depend on climate, temperature and local conditions.

The particulate solids in raw animal slurries (especially after drying in high outdoor temperatures) may either sink or float as the slurry stands in tanks (or pipes). Fibrous solids tend to form a "screen" and allow only the supernatant liquids to flow. Floating fibrous solids can dry out, to give a solid crust on slurries, which can be difficult to break up. This is especially true when low concentrations (4 - 6% solids) are used for digestion. Settled solids can block pipes and pumps that are taking slurry from the bottom of the tanks. After some time, settled solids can become very difficult to break up. Animal slurries have been mentioned, but solids settlement or flotation can occur in many industrial wastewaters, such as those containing starch granules, vegetable peels, yeasts from brewing, fragments of seeds from oil-extraction plant, slaughterhouse wastewater, etc. The problems can occur in pipes and channels or in feed tanks. Mixing of solids and liquid can sometimes be obtained by an intermittent rapid flow, as when sluices are opened to empty a slurry channel, or when slurry is circulated over the top of the tank. Pumps can be designed to recirculate liquid in a tank for some time, before switching to pumping the liquid out to another tank, or to a digester, or to macerate the materials for a short period before pumping. Paddle, Archimedean-screw or other mixers, or independent recirculation pumps may also be used. A number of macerators can be purchased cheaply from several companies. In the case of cotton stalks as with vegetable wastes, energy crops, fresh or ensiled, a stable sludge may be obtained by macerating the vegetation in water. Maceration may also prevent pipe and pump blockages by long-fibre vegetable material in animal excrete and other slurries. These methods attempt to keep the whole digester feedstock, or effluent in some cases, as a "homogeneous" suspension of solids. On the other hand it may in some cases be better to remove some, or all, of the solids. Slurries and sludges can be treated only in stirred-tank digesters; removal of suspended solids may allow the resulting liquid to be treated in some simpler digestion system. Some simple screens are used in cow barns to separate solids and fibres from the liquids and treat each of them differently. Settling tanks for grit are sometimes essential.

Solid Feedstocks. Because of the nature of the solid waste it may not be possible to use a continuous feed. In this case, batch digestion may be used. Batch digesters are of the single-stage, tank type or plug-flow, but may have either a liquid or a solid feedstock. Most batch digesters are designed for solid feeds, the type of digester being dictated by the above-mentioned difficulties of continuously feeding a solid substrate. The feed has to be mixed with an inoculum of the previously digested material, in order to start off the digestion process. Gas production then proceeds at an increasing and then decreasing rate, as the substrates are used up. To deal with a continuous supply of solid feed, and to produce a continuous supply of gas, a number of batch digesters must be run in rotation, one producing maximum gas, while others are in the starting-up or declining phases of gas production. This was very nicely shown in the MAYA farm in the Philippines.

Manures usually have a total solid (TS) content of about 8 16%. In most cases, especially when slurries are not "homogeneous", this concentration is too high for pumping and piping, and the TS is usually brought down to 6 - 8% by dilution with water, to make a pumpable slurry. Some vegetable matter (eg. seaweeds) may have such a high water content that the macerated material is of this order of TS, as well as pig manure and cow manure in "flashed system". However, some wastes are "solids" materials with more than 20% TS, and although containing water, are not pumpable. Examples of solid wastes are poultry excreta, with or without litter, and cattle excreta, with straw or other bedding. Many plant waste" and terrestrial plants grown as energy crops will give a digester feed of 25 - 30% TS. Unless this type of feedstock i" suspended in water and macerated, to produce a low- solids slurry it can be easily handled only in "solid-state" digesters (Jewell 1980). These are run batch-wise and are loaded and unloaded by grabs or tractor shovels which may feed the digester directly or indirectly. Even in this case, some chopping of large pieces of vegetation may be needed, but on the whole, solid-state digesters have few physical problems with either feed or effluent. The system is very similar to silage making but is rarely used for biogas production.

Digestion of solid materials has to be a continuous process, though the feedstock itself may be produced intermittently, as with some energy crops and crop residues, for instance. Continuous use of any digester, as with other chemical or power plants, is dictated by economic considerations of obtaining the best return on capital from the gas or other products of digestion (except in pollution control cases). Sometimes the digester will work only when the factory is operating. However, though the running of a digester may be intermittent, in the broader sense, running is as far as possible continuous during the period of use. Sayed et al. (1987) ran a UASB digester on slaughterhouse wastewaters. Feed was continuous during the week but stopped at weekends when the slaughterhouse was not in operation. Analysis showed that this weekend of pause in loading was beneficial, in that it allowed accumulated solids in the digester to be degraded to gas. Marchaim et al. (1991) found a similar situation in treating high solid concentration of slaughterhouse wastes.

Although a solid feedstock will not be pumpable in the usual sense, it may be possible to move it through pipes by conveyors. Thus, if a suitable digester can be designed, a continuous-flow digestion, similar to the stirred-tank for liquids, can be run. A typical digester is used for solid municipal refuse (de Baere et al. 1986).

Physical Pretreatment of Slurries and Solid Feedstocks Maceration of feeds has been mentioned above. With drier materials, such as straws, either alone or as constituents of a mixture (e.g. animal excrete and bedding), maceration reduces particle size to prevent physical obstruction of pipes and pumps by the fibres, and it also increases surface area available for microbial attack, and thus speeds up the digestion process. It was shown that lignin cellulose and hemi-cellulose which are almost non-biodegradable in ordinary systems can be degraded to an significant degree after maceration (Marchaim 1983). With fresher vegetable matter, such as seaweeds and water hyacinths, the cell walls of the algae or plants are only slowly degradable, and maceration frees soluble cell-contents, which provide the bulk of the digestible substrates, and also damages the cell walls to make them more susceptible to digestion. This type of feedstock should be relatively free of stones and other debris, or must be washed free of such material, so some type of rotary chopper can be used to reduce the particle size.

Municipal garbage and similar solids feeds may be mechanically, magnetically or hand-sorted, to remove large glass, metal and other non-biodegradable materials, and then macerated in an internally-toothed drum or other apparatus which will shred the waste, but will not be affected by stones or other hard debris. This type of shredder is not suitable for liquid slurries or slurries containing stones, lumps of wood, plastic sheets, etc. (eg. farm slurries) that may damage the choppers. This sort of operation is not dealt with in this review.

Separation of solids from farm slurries, particularly cattle wastes, has been advocated by many companies and farmers. Separation is done to remove the long plant fibres and cattle hairs, thus decreasing the possibility of pump and pipe blockages, and making a more homogeneous slurry. The large fibres removed are generally poorly biodegradable, and produce gas only at long retention times. The remaining particulate matter can be digested at short retention times, thus allowing the use of smaller digesters and lower capital costs. There will be some overall loss of gas production from a particular volume of original slurry, but this can be compensated for by the smaller digester and lower cost and more trouble-free pumping and mixing (Pain et al. 1984). The digested slurry obtained has a different structure and value than the "whole material" digested slurry. The solids removed could be digested in a long retention-time batch digester, but are more often aerobically composted, as they are relatively dry, stackable solids. Removal of most of the particulate matter by mechanical mean" or by gravity settlement can be performed. This leaves a wastewater of only 1 - 2% TS, most of which is dissolved. This waste can then be treated in a filter or UASB digester, at a relatively short HRT. Obviously, in this case, most of the gas production potential of the original slurry is lost. The anaerobic filter then becomes merely a method of reducing pollution in the lagoon overflow. The solids in the lagoon digest slowly, and gas escapes to the atmosphere, but it can be collected and used as shown by Balsari and Bozza (1987). Solids may also be allowed to settle in a lagoon or tank and degrade to acids. The tank then becomes part of a two-phase digestion in which the supernatant liquid is converted to gas in an anaerobic filter, (Colleran et al. 1982). There is still a liquid sludge in the tank to dispose of, which may be only partly stabilized.

Separation of solids from animal wastes allows the liquid to be treated in an anaerobic filter, but it removes much biodegradable material. Separation may also be used to remove undegradable material. For instance, Hobson (1987) found that yeast in a whisky distillery spent wash (the residue from the stills) was essentially non-digestible, but its presence caused mechanical problems in digesters. Allowing the yeast to settle out produced a relatively clear liquid which could be treated in filter or UASB digesters. No potential gas production was lost by this separation, and the yeast could be dried for use as cattle feed. Other examples are the separation of solids from palm-oil materials. There is thus good reason for separation and the use of a retained -biomass digester instead of a stirred-tank model. However, mechanical separation is costly in term of capital and running, and separators are not trouble-free. If separation also results in loss of biogas which could be used, then careful consideration must be given to whether the benefits of separation (in the use of a simpler digester for instance) outweigh these increased costs.

Chemical and biochemical treatments of feedstocks

Physical treatments, such as maceration of feedstocks, can increase the rate and extent of bacterial digestion and thus the gas production, but they do not essentially change the chemical composition of the feed. The reason for changing the composition of feedstocks is to break down fibrous materials and produce compounds which are better substrates for microbial growth than the original material. For example, plant materials, either residues in animal excrete or vegetation in energy-crop or plant- waste feeds, can be partly degraded by chemical treatment. Many types of chemical, physical and biological treatments have been applied to plant materials to break down the lignin and make the vegetation more digestible by ruminants (Tagari 1978) and similar treatments could be applied to anaerobic digester feedstocks. The most successful treatments have been those where alkalis, usually ammonia or NaOH, are allowed to react and hydrolyze the solid vegetation for some days at ambient temperatures. Ammonia is used to raise the nitrogen content and decrease the C:N ratio. This type of treatment could be applied to solid feedstocks, such as straw or other plant material. With slurries, it would be difficult to obtain the necessary concentration of chemical (equivalent to a few percent by weight of dry animal feed) to raise the pH sufficiently for alkaline degradation of the lignin to take place, without producing sodium or ammonia concentrations, or an irreversibly high pH, which could inhibit the subsequent anaerobic digestion.

The examples show that some changes may need to be made to digester feedstocks to obtain optimal digestion. These changes may involve the addition of specific chemicals to the feed, but the same changes might be brought about by the addition of an undefined material. Such material may itself be a potential digester feedstock, and mixing may allow better digestion of two or more components of the waste streams. Reactions, such as fermentation of sugars, may go on during the collection process and some alkali or other treatment may be needed before the waste is fed to the digester. Poultry manure, which is rich in nitrogen, can balance a material with low nitrogen content, such as vegetable wastes.

Pretreatment of a feedstock may be inadvertent, in that changes may take place in the feed while it is stored, prior to delivery to the digester. The breakdown on storage of a primary feedstock may be allowed or encouraged, if it results in the production of a substance which itself is a digester substrate. This is the basis for the two-phase digestion of farm wastes. Stopping the natural fermentation of a feed containing easily degraded material like sugars may be difficult, impossible or too expensive, so it may be better to encourage this fermentation to acids to proceed to completion, perhaps by the addition of alkali, to prevent the inhibition of fermentation by low pH. The acid containing solution can then be treated in a retained-biomass digester to give methane. Ensilage is also a natural acid fermentation process, and ensilage of vegetation to store seasonally-produced plants for digestion has been practiced for some years (Stewart 1981). Ensilage produces acids which prevent subsequent deleterious microbial growth, and which can also make the plant fibres more digestible. The acids, along with the solids, can be degraded in the digester after storage. Coble and Egg (1987) ensiled sweet sorghum, which has a high sugar content, but instead of allowing the reaction to stop naturally when the pH had fallen (as happens normally in silage making), they encouraged continual fermentation of the sorghum by connecting the silo to an anaerobic filter, and circulating the silo leachate (plus added water) through the filter and back through the silo.

Digester effluent

As with feedstocks, there should be fewer problems in removing effluents of wastewater digestion from the digester: a U- shape pipe, to retain gas, and gravity flow should suffice. With slurries, weir systems and gravity flow are often sufficient, but if the slurry is thick, or if it has to be moved some distance from the digester for storage or treatment, a pumped output may be better than gravity flow. In general, any treatment which helps the handling of feed will facilitate handling of effluent. Treatment to be applied to effluents depends on many factors. If the effluent is to be used as fertilizer, storage until land and crop conditions allow spreading may be all that is needed. Storage will permit some residual digestion and further reduction of pollutants to take place. With contact type digesters, separation and recycling of digester bacteria is necessary. Separation and recycling of residual solids may help in digestion of recalcitrant materials at a low HRT in a stirred-tank digester. Separation of solids may also be practiced to obtain a solid which can be composted, to be used as soil conditioner, and a liquid which can be used as fertilizer, irrigation water, or recycled to dilute a thick feedstock.

Separation of solids will be necessary if the BOD of the effluent is to be further reduced for river discharge. In the case of a retained-biomass digester, the only separation required is gravity settling of the relatively small amount of biomass which breaks off from the biomass in the digester. In the case of a slurry with undigested solids from a stirred-tank digester, mechanical separation of fibres and sands are required. The liquids can be aerobically treated and filtered, or further treated, as required by the discharge conditions.

Depending on climate, land available, etc., separated liquids can be used for growing algae, plants, or fish (Marchaim 1983). This will further purify the water and produce an added-value crop. As with the feeds, mechanical separation will produce stackable solids, which can be comported or used immediately as fertilizer or perhaps soil conditioner. Depending on circumstances, it may be possible to use separated effluent solids as protein components for farm animal diets. Since the material has been digested, it is unlikely to contain any substances which are immediately toxic to microbial, plant or animal life. Heavy metals which have been immobilized by precipitation in the digester can be slowly released in fertilized ground or in animal guts, if the effluent is being used as a feedstuff. Such problems have to be borne in mind when the destination of the effluent is being considered.

Control device in an anaerobic digestion process

Anaerobic digestion of organic materials is to be one of the more accepted biomass processes, because it has been in use by numerous municipalities and companies for many years for waste treatment. Over the last two decades, it has been considered and developed for the production of energy; however, the acceptance of the technology for energy production has been limited. The limitations and problems of the process have been the subject of considerable research and development throughout the world; one key topic is the inhibition of the digestion process and its causes.

Ideal indicators for process inhibition should be capable of measuring the progress of sludge digestion, and signal impending upsets before they occur. Common indicators, such as volatile fatty acids, gas composition and pH, are useful for monitoring gradual changes, but do not directly reflect the current metabolic status of the active organisms in the system. They are generally useful for detecting process upsets once they are underway, but in most instances are not adequate to avoid system failure due to difficulties, such as gradual organic or hydraulic overloads.

The mechanism of the rate-limiting step in methane fermentation can be and has been debated, but clearly involves the degradation of volatile fatty acids during methanogenesis, since these acids begin to accumulate in digesters stressed by high organic loading rates and/or short retention times and/or inhibitors (Mackie and Bryant, 1981; Ashley and Hurst, 1981; McInerney et al., 1981). The importance of short-chain fatty acids and alcohols as intermediate metabolites during anaerobic digestion has been well recognized (Smith and Mah 1978). The further degradation of these intermediates relies on dehydrogenation reactions. The energetics of these reactions are only favourable when the concentration of hydrogen is kept very low. The microorganisms that catalyze these dehydrations are considered to be obligate syntrophs, able to grow only in the presence of hydrogen consuming organisms (Bryant et al. 1979, Boone and Bryant 1980, McInerny and Bryant 1981, McInerny et al. 1981). Hydrogen concentrations in anaerobic digestion systems range between 5 - 10 nM when hydrogen-consuming methanogens are active (Poels et al. 1985, Archer et al. 1986, Hickey et al. 1987). An understanding of the close relationship between hydrogen producing and hydrogen consuming organisms has led to the concept of interspecies hydrogen transfer (Bryant et al. 1967, Reddy et al. 1972, Iannotti et al. 1973, Scheifinger et al. 1975, Wolin 1982).

Increases in the concentration of hydrogen would be expected to inhibit the synthropic hydrogen producing partner. For instance, addition of hydrogen to a co-culture of a butyrate- degrader and a methanogen led to the inhibition of butyrate degradation (Ahring and Westermann 1987a, Ahring and Westermann 1987b). It has been suggested that the kinetics of hydrogen consumption controls the bioenergetics and rate of fatty acids oxidation (Dwyer et al. 1988) in defined co-cultures. The exact relationship between hydrogen and fatty acids oxidation remains to be elucidated, but is of critical importance to development of a feedback system based on hydrogen, acetate and propionate monitoring.

This three-step microbial process can work integratively in a digester (the CSTR systems) or can work separately in two (or three) stage digesters, in which the acidogenesis takes place in one digester in its optimal condition, and the effluent is then transferred to the methanogenesis step in another reactor (Cohen 1980; Pipyn and Verstraete 1981). In most industrial uses of anaerobic digestion the CSTR system is used; and a mixed population of acidogenic and methanogenic bacteria is therefore present, such that each probably is not operating in its optimal environment.

It has been shown by several groups that the relationship between methanogenic and non-methanogenic bacteria during the anaerobic digestion i" of great importance, and various growth conditions affect the methane production. For example, when methanogenesis was inhibited, a high volatile fatty acids concentration was observed in the effluent (Sorensen et al. 1981; Ashley and Hurst 1981), the volatile fatty acids being acetic, propionic, butyric acids and others. In many cases described in the literature, unbalanced digestion results in a relatively high concentration of propionic acid, and pH drops below the preferred range. The control of pH is difficult to achieve because of this build-up of volatile fatty acids and, hence, pH control has been only moderately successful.

One of the important ways to control the microbiological process of anaerobic digestion is to control the organic loading to the system (Cohen, 1980). Measurements of volatile acids in good, continuously operating systems show variable levels of acetic acid (according to specific conditions), but very low concentration of propionic acid. It was also shown that during inhibition the level of propionic acid rises, which suggests a shift in microorganism activity. Cohen (1980), working with a physical separation of the acidogens and methanogens by two reactors (each adjusted for optimal conditions) showed that when the first stage digester is overloaded, considerable amounts of propionate and acetate are formed; and, although acetate disappeared rapidly after cessation of the feed, no turndown of propionate was observed. In some cases, accumulation of hydrogen has been noted when inhibition of methanogenesis takes place, while volatile fatty acids also accumulate.

As can be seen from the different biochemical reactions, the important change is in the hydrogen pathway. While hydrogen is generated with acetic acid in the regular pathway of glucose breakdown, hydrogen is consumed in order to produce propionic acid from glucose. The propionate accumulation is probably not a regular intermediate during good steady-state digestion of glucose. Thus, one might view the propionic pathway as a "hydrogen-sink" .

Hydrogen partial pressure (PH₂) exerts significant control on bacterial populations and their Interactions in methanogenic processes. McCarty and Smith (1986), and others, showed that the microbiological population does not appear to utilize the accumulated propionic acid efficiently while rapidly assimilating acetic acid. This leads to a hypothesis that control of an anaerobic digester should strive to prevent the production of propionic acid, because of the microbiological and biochemical shift it implies. Further, the above reactions imply that the key route for achieving this control or stabilization is to alter (decrease) the organic loading when propionic acid accumulation is observed.

Process control

Several concepts of control systems are known and used, especially in the chemical industry, such as those described by McClain and Goswami (1979) (closed loop system, pacing system, and ratio control system). There is a need to improve existing anaerobic digester control systems and develop better ones, incorporating such concepts applied in other industries.

This need was amply illustrated by the number of times that the topic was questioned and discussed at various International symposiums on anaerobic digestion - AD83, AD85, AD88. The general consensus seemed to be that the usual control parameters (e.g. pH control; Schaffer and Casciano 1979) were not sufficient, or the measurement was not reliable. It must be noted that none of the usual analysis/control schemes provide a direct link to the biochemistry related to the microbiology of the digestion system. Thus, what appears to be needed is a control system which is more closely related to the biochemistry of the digester, and which employs reliable analytical approach and equipment.

Since 1985, studies have been conducted by Hickey and colleagues (1988) to evaluate the efficiency of hydrogen and CO as a means of monitoring anaerobic systems, under steady state conditions, in response to organic overloads and in response to toxic or inhibitory shocks induced through the application of organic and inorganic toxicants. The long range primary goal of this research was to use this information data base to assist in developing more sensitive and effective process control strategies, that will minimize the time interval between the impending upset and its actual occurrence. This will allow more time for remedial measures, and help to eliminate or alleviate the severity of an oncoming upset. It was found that the impact that the toxins (and organic overloads) had on the system were concurrently assessed by more conventional parameters (gas production, methane and carbon dioxide composition, VFA and pH), to allow a comparison of the various parameters.

A large data base on the response of the anaerobic sludge digestion systems to overloading and toxic inputs has been generated. Results have been published on the effect of organic toxicants (Hickey et al. 1988) in terms of hydrogen response. The conventional parameters show a slow deterioration in process performance. The VFA/TA ratio was the first to register a possible upset condition. Gas production and carbon dioxide/methane content did not show any definite indication of upset until a few days later. Hydrogen and CO, however, demonstrated a significant initial response to increase in organic loading. A strong correlation between CO and acetate is apparent from the data. It also appears that there may be some correlation between hydrogen concentration and gas production rate, as has been suggested (Mosey 1983).

To summarize world efforts up to the present time, monitoring intermediate trace gases and volatile fatty acids can provide some measure of the metabolic status of an anaerobic system. In contrast to liquid phase sampling, gas analysis. is amenable to real time data acquisition. Using a tiered experimental plan, the response of systems due to toxic upsets and variations in hydraulic and organic loading has been studied along with more conventional parameters of digester stability. Based on these studies, it appears that monitoring both hydrogen and CO together, and volatile fatty acids, may lend significant insight into the metabolic status of the digestion process and has the potential to indicate process upsets on a real-time basis. The exact relationship between hydrogen and fatty acids oxidation remains to be elucidated, but is of critical importance to development of a feedback system based on hydrogen, acetate and propionate monitoring. It has been shown by several groups that the inter-relationship between methanogenic and non-methanogenic bacteria during anaerobic digestion is of great importance, and various growth conditions affect methane production. For example, when methanogenesis was inhibited, a high volatile fatty acids concentration was observed in the effluent (Sorensen et al. 1981; Ashley and Hurst 1981), namely acetic, propionic, butyric acids and others. In many cases described in the literature, unbalanced digestion results in a relatively high concentration of propionic acid; and pH drops below the preferred range. The control of pH is difficult to achieve because of this build up of volatile fatty acids, and hence, pH control has been only moderately successful.

Experiments were performed by Marchaim and Krause (1991) in order to examine the possibility of controlling the anaerobic system by monitoring the acetic to propionic acids ratio. In the Marchaim and Krause experiment, the system consisted of an anaerobic digestion system with a steady pH level (approximately pH 7). Changes in the rate of feeding with glucose, were made in several digesters, in order to compare the ratio of propionic to acetic acids in the overloaded digesters, which were operated under identical conditions.

The fact that the ratio of propionic to acetic acid showed an increase immediately after raising the concentration of feeding, and prior to any changes in biogas composition, suggests that the ratio of propionic to acetic acid in an anaerobic digestion system is a satisfactory real-time indicator for the beginning of organic overloading.

Contents - Previous - Next