Organic Waste for Biological Treatment
INTRODUCTION
The diverse metabolic and physiological characteristics of microorganisms, in addition to their ability to thrive in a variety of environments, may be exploited for the purposes of environmental remediation and waste treatment Axenic cultures or mixed populations of microorganisms with the ability to degrade or mineralize hazardous materials can form the basis of a bioprocess for the treatment of organic hazardous waste or serve as a refinement of many conventional waste treatment processes. Research has progressed from “bench scale studies” to field and treatability studies for this remediation of contaminated sites. In many instances, bioremediation is being actively pursued as the preferred treatment option.
PHYSIOLOGICAL BASIS FOR HAZARDOUS ORGANIC
COMPOUND DEGRADATION
A survey of the literature reveals a wealth of information concerning the degradation of a diverse array of organic compounds by a variety of micro-organisms as well as reports of the degradation of organic chemicals in the environment by a combination of biotic and abiotic processes. Advancement in biologically based treatment strategies for hazardous waste treatment and environmental remediation is, in part, dependent upon understanding the mechanisms by which organisms degrade organic materials. The intent of this chapter is not an exhaustive review of this literature, but rather a discussion of the features common to the microbial degradation of organic compounds and their implication for hazardous waste treatment and environmental remediation.
Miciobial transformations applicable to waste treatment are those reactions that microorganisms mediate to satisfy nutritional requirements, satisfy energy requirements, detoxify their immediate environment, or are the indirect or unexpected result of metabolic processes or the physical or chemical characteristics of the microbial cell. While many transformations applicable to waste treatment are of a direct benefit to the microorganisms, some transformations are the result of fortuitous reactions that do not offer any advantage to the microorganisms.
In order to satisfy their nutritional and energy requirements, heterotrophic microorganisms metabolize a variety of organic compounds. Anabolic metabolism results in the synthesis of a diverse array of monomers, polymers, and complex macromolecules from simple precursor molecules. Often the precursor molecules are derived from the degradation or catabolism of other, frequently more complex, organic molecules. Important to the microbial degradation of hazardous organic materials are the catabolic or degradative capabilities that enable microorganisms to degrade these compounds and utilize them as suitable substrates for growth.
The transformations resulting in the degradation of organic materials can be classified into two broad categories: mineralization and cometabolism. Whether a compound is mineralized or cometabolized has implications for the development of a waste treatment process or for environmental remediation. Mineralization is the complete conversion of organic materials to inorganic products. Mineralization is a growth-linked process that involves the central catabolic and anabolic pathways of a microorganism. A compound that is mineralized serves as the growth substrate and energy source for the microorganism. In general, only a portion of the organic compound is incorporated into cell material with the remainder forming metabolic by-products such as CO2 and H2O. Mineralization can also occur by the combined activities of a microbial consortia. Interdependent members of the community are required to effect the complete conversion of the hazardous material to satisfy growth requirements.
Cometabolism is the degradation of organic compounds usually via nonspecific enzymatically mediated transformations. In contrast to mineralization, cometabolism does not result in the increase in cell biomass or energy. Consequently, the ability to cometabolize a compound is not a benefit to the microorganism. In fact, another substrate is necessary in order to satisfy growth and energy requirements of the cell. Typically, cometabolism results in the modification or transformation of the organic material and does not result in the complete destruction of the molecule. While cometabolism can result in the complete destruction of an organic molecule, the accumulation of potentially toxic intermediates in the environment can occur.
ANAEROBIC VERSUS AEROBIC TRANSFORMATIONS
Suitable electron donors and electron acceptors are required for the energy yielding reactions of the cell. Microorganisms that utilize organic substrates as the source of carbon for growth typically utilize the organic substrate as the electron donor (energy source) as well. Depending on the microorganism, a variety of inorganic species and organic molecules can serve as electron acceptors. Aerobic microorganisms utilize oxygen as the terminal electron acceptor and are limited to environments that contain sufficient amounts of oxygen. Anaerobic microorganisms are those microorganisms that do not utilize oxygen as the terminal electron acceptor. Anaerobic microorganisms use arrange of electron acceptors, which, depending on the redox conditions and availability, include nitrate, iron, manganese, sulfate, and CO2.
Hazardous organic materials can be biodegraded under aerobic or anaerobic conditions. The type and form of the organic hazardous material, the catabolic capabilities of the microorganisms, and the availability of nutrients and electron acceptors may dictate whether aerobic or anaerobic degradation is the preferred treatment option. In cases where there is little or no oxygen or there is an abundance of an alternative electron acceptor such as nitrate or sulfate, degradation of hazardous organic materials by anaerobic microorganisms should be considered. The degradation of organics by anaerobic bacteria is of importance when remediating anoxic environments such as saturated subsurface soils, landfills, lagoons, and some groundwaters.
EXAMPLES OF HAZARDOUS WASTE AMENABLE TO BIOLOGICAL TREATMENT
Organic compounds that can be metabolized by microorganisms are amenable to treatment by waste treatment bioprocesses. The mineralization of organic chemicals by microorganisms or microbial consortia that utilize these compounds for their carbon and energy source assures the complete destruction of these hazardous materials. Hence, mineralization of organic materials is preferred for the microbial treatment of organic waste or environmental remediation as it potentially offers a permanent solution. Microbially based treatment processes are suitable for concentrated wastes as well as dilute aqueous based wastes. Thus, microbial degradation is perceived as a low cost alternative to incineration of concentrated waste forms as well as other oxidative treatments such as the photooxidation of dilute waste streams.
Partial modification of hazardous organics is not considered an environmentally sound treatment option unless another microbial population is present to complete the destruction. For example, partial degradation of trichloroethylene can lead to the accumulation of chlorinated intermediates such as dichloroethylene or vinyl chloride. In essence, the problem of trichloroethylene contamination can be exchanged for dichloroethylene or vinyl chloride contamination unless another population is present that can degrade these toxic intermediates. For this reason, transformations that lead to the mineralization of organic materials by axenic cultures or microbial consortia will be emphasized in the remainder of this chapter.
Whether a compound is mineralized anaerobically or aerobically, a review of the degradation pathways reveals the fact that many compounds are degraded via common metabolic intermediates. It is the peripheral transformations that transform diverse compounds to common intermediates. Within a class of compounds, such as aromatic hydrocarbons, features common to the catabolism of the seemingly different compounds are readily apparent
The description of pathways applicable to hazardous waste treatment will begin by a discussion of the degradation of aliphatic and aromatic compounds found in petroleum. The section will continue with a discussion of the transformation of halogenated compounds. Halogenated compounds, once dehalogenated, are degraded by the pathways involved with aromatic or aliphatic hydrocarbon degradation.
Fermentation of Fish Waste
Fish protein is used extensively in poultry and swine production as a source of high-quality protein. Because the production of fish meal has leveled off in the past years and available fish stocks are not sufficient to sustain increased production levels, supplies have not been adequate. Limited supplies and the supplies and the demand for fish meal as a source of human food has made it necessary to look for alternative sources offish protein for animal feed. Waste products from fish-processing operations and the use of underutilized species of fish to produce fish protein for animal feed have been studied. The use of fermentation or acid addition to produce fish silage has
been studied as a way of using these sources to produce fish protein for use in animal feed. Fish silage has been defined as a liquid product from whole or parts of fish to which no materials are added other than acid, and in which liquefaction of the fish mass is carried out by enzymes present in the fish. Fish silage has become of interest to the poultry industry as a feed ingredient and this interest has been directed toward methods of production using either direct acid addition or biological fermentation. Most of the information is available on the direct acid addition method with comparatively little information on the biological fermentation method. Biological fermentation has been used to produce a fish silage product by using microorganisms to produce the acid. Stanton investigated the use of lactic acid fermentation at low salt concentration to produce fish silage under South-East Asian conditions. They used a local source of carbohydrate, tapioca starch, as an energy source. Results showed that fish silage can be successfully produced using a 1:1 fish/starch ratio and adding ragi and lactic acid starter cultures. March (1962) indicated that the application of molasses and inoculation with lactic acid bacteria was the basis of a patented Danish method. The effect of fish silage (14-60 g hen”’day’1) on the quality of hen’s eggs and broiler meat was studied. He found that the eggshells from hens fed fish silage were thicker and denser than shells from hens fed fish meal. He theorized that the lactic acid present in the fish silage apparently facilitated the uptake of calcium. The yolks of eggs from hens fed with fish silage were less yellow and had a higher iodine value than yolks from hens fed fish meal. Stability of the egg foam did not differ between the two treatments. Feeding 40 g of wet fish silage per hen per day did not impair the taste or flavor of the eggs. Feeding broilers at the same level resulted in off-taste and flavor within 7 weeks, but when silage was removed 1 week prior to slaughter no off-taste and flavor were observed. It was concluded that fish silage can be incorporated into the diets of both broilers and hens if precautions are observed to prevent off-taste or
flavor. Norwegian research has demonstrated that fish silage is equal to herring meal for growth and egg production in hens: superior hatchability was also reported. Laying birds may be fed about 250g of fish silage per 12 birds without producing tainted eggs, and breeding fowl may be fed 50 % more than this.
The objective of this research was to provide information to facilitate production of fish silage from waste and underutilized fish by biological fermentation for potential use in animal and poultry feed. To provide this information, levels of carbohydrate necessary for successful fermentation, and the effect(s) on the silage product of using whole fish or offal, were evaluated. The effects of preheating the fish, yeast and mold contamination, fermentation temperature and inoculum size on the fermentation process were also investigated.
METHODS
Fish
White perch were used in this study to provide a low-oil fish that would be representative of some of the underutilized species and would provide material that was representative of waste material from commercial processing operations.
Sample preparation
Whole fish and viscera and heads were required for the studies. Whole fish were defined as the entire fish as they came from the water. Viscera and heads were obtained by making a longitudinal cut from the back of the head to the anus. Whole fish and viscera and head samples were ground separately in a meat grinder using plates with 5mm, 2mm and 1.5 mm diameter pores. The samples were ground through the plates in order of size starting with the largest pore size.
Inoculum
Lactobacillus plantarum was-used to inoculate the fish for fermentation because it had been proved to ferment fish efficiently under the conditions used in the present experiments. The organism was obtained from the culture collection of the Food Science Program at the University of Maryland and maintained in litmus milk at 4°C. Prior to use, it was activated by three successive transfers into MRS broth. Inoculation was made from a culture grown for 18 h.
Analytical methods
5 g of fish were mixed with distilled water to a volume of 50 ml, centrifuged at 8000 rev min–1 for 10 min using a Sorvall RC-2 refrigerated centrifuge, and pH was measured on the supernatant.
Titratable acidity was measured by titrating 10ml of the supernatant obtained as described for pH determination with 0.1 N NaOH using phenolphthalein as the indicator. Acidity was expressed as % lactic acid by applying the following formula:
% Lactic acid=(Vol. NaOH) (N of NaOH) (eq. wt of lactic acid) (100)
1000 (g of sample)
Percent developed acidity was determined by subtracting the % titratable acidity of fish before fermentation from the % titratable acidity of the fermented sample.
Soluble nitrogen was determined according to the method of Barbour. Moisture, fat, protein and ash were measured using AOAC methods.
Microbiological examination
The sample was prepared by homogenizing 10 g of fermented fish in 90 ml of peptone water (0.1%) for 2 min, and dilutions of 10–2 to 10–8 were prepared. Total viable counts were determined using the procedure of Gilliland and total mold and yeast counts were determined as outlined by Koburger.
Statistical analysis
Analysis of variance was calculated using the computer program BMDP2V and BMD1R at the University of Maryland Computer Science Center.
Experiment 1
100g of either ground fish or ground viscera and heads were placed in glass, screw-capped jars. To each jar 2,3,4 or 5 % lactose was added, then the sample was inoculated with 1 ml of a culture of L. plantarum grown for 18 h in MRS broth. The preparation was mixed thoroughly, sealed and incubated in a water bath at 35 °C. After 2 and 7 days, 5 g of fish were aseptically removed and mixed with 45 ml of distilled water. Titratable acidity and pH were measured on each sample.
Samples were removed from the whole fish and the viscera and head samples prior to fermentation for proximate composition analysis.
The lactose used in this and subsequent experiments was obtained
from Foremost Foods, Inc.
Palm Oil Mill Effluent Disposal on Land
Legislation for the control of effluent disposal from palm oil mills in Malaysia is in the process of being brought into force. The Government has specifically envisaged land application as one disposal alternative, either after digestion has reduced BOD levels to 5000 mg litre–1 or raw in individually approved schemes. The effluent does not contain toxic elements in significant quantity and in its raw state it is high in organic matter. Such effluents are increasingly being applied on the land in various parts of the world, with good agricultural benefits. Indeed, there is a growing feeling that simply to treat and then discharge them is wasteful of a valuable resource.
Our intention in this paper is to report trials on the application of palm oil mill effluents to crop land. Application of such effluents in tropical conditions is a new field of study, with little information so far published, so that the techniques of assessment are necessarily being developed along with the experiments. Further, long-term effects must be considered. To this extent, publication now may be considered premature, but the nature of the problem is so pressing, and results to date of such promise, that it is felt worth while to describe the work so that others may carry out similar trials to increase the range of circumstances involved and ideas brought to bear in the development.
Palm oil mill effluent (POME) consists of fruit constituents plus, at most, tiny amounts of material eroded or otherwise picked up from processing machinery, dissolved or suspended in processing water. It has three main components. The principal one is sludge from the clarification tanks (POMS) which has the highest solids content. The others are condensate from the steriliser and water from the hydrocyclone. The output is considerable, at about 60% of the total fruit bunch weight processed (45, 10 and 5% in the three mentioned components, respectively). In our work, both POMS and the more dilute POME were used. The suspended material is colloidal and contains carbohydrates with oil and other organic and inorganic solids. The effluent has various salts and ions in solution and extremely high BOD and COD values. A typical analysis is given in Table 1. In addition, there will be a volume of water, used for washing down and general purposes, which will vary in quantity between mills but will normally be around 10-20% of the fruit volume input. Much of this will flow at different times than the POME whilst the factory is not operating, but it will have to be taken into account in any application project.
Early efforts to apply POME to land were unsatisfactory. Vegetation was killed on contact and, where it was applied in quantity, accumulation of a gel and clogging of the soil occurred so that percolation slowed down drastically, leading to stagnation of rainwater and run-off. This led to odour and difficult ground conditions, making working intolerable. Furrows became impervious and effluent simply ran along them rather than gradually percolating away as in irrigation with water or solutions.
TABLE 1
CONSTITUENTS OF POME BASED ON FIVE SUCCESSIVE DAILY SAMPLES FROM THREE SEPARATE MILLS
Concentration Mean (mg litre–1) Mean quantity Kg
in one
Mean Range hectare-cm
(100000 litres)
BOD 20800 17000-26700 2100
COD 57900 42900-88250 5800
Oil (vegetable) 6300 4400-8000 628
Nitrogen (total) 643 500-800 64
Phosphorus 110 94-131 11
Potassium 1620 1281-1928 162
Magnesium 295 254-344 30
Calcium 315 276-405 32
Manganese 3.3 2.1-4.4 0.4
Iron 117 75-164 11
Zinc 1.5 1.2-1.8 0.14
Copper 1.2 0.8-1.6 0.09
Chromium 0.17 0.05-0.43 0.01
Cobalt 0.05 0.04-0.06 0.004
Cadmium 0.01 0.01-0.02 0.001
Total solid 38900 29600-55400 3900
Suspended solids 18500 14100-26400 1840
Dissolved solids 20400 15500-29000 2000
aMean quantities of each constituent which would be applied to land at the effluent dosage rate shown.
In spite of these indications and encouraged by the success in applying ubber effluent to land we continued tests on land application. We had noted, when attempting to filter oil palm effluent, that, with small quantities, the liquid component flows through the filter paper to leave a layer of solids. In larger quantities this solid causes blocking and stops filtration. It appeared to us by analogy that a similar effect might occur on land, and this led us to re-examine the problem using a ‘little-at-a-time’ technique.
METHODS AND RESULTS
Infiltration tests
The capacities of soils for infiltration of water were compared in the field using a concentric ring infiltrometer. This comprises a 36 cm diameter metal cylinder, 23 cm deep, which is hammered into the soil to a depth of 8 cm, with a 16 cm diameter cylinder similarly placed in the centre. Water is poured into the cylinders ad lib for a 20 min stabilisation period, after which the centre cylinder is kept topped up and the rate of outflow of water is measured at intervals, for about 1½ h.
Small-scale trials of the general effects of application of sludge (POMS) to land Trial1:
A small-plot trial was devised to assess the effect of applying the filtrate after filtering the POMS. However, none of the filtration methods (rotary drum screening, fibre, etc.) could consistently remove solids when maximum water economy was practised in the factory operation and no account of these treatments is given. Raw sludge was incorporated in the trial as a control and it was this that proved interesting.
The POMS was applied to 1.5 m square plots,cleared of vegetation and levelled, comparing a total volume equivalent to 5 cm per month, applied twice a week, with water applied at the same rate. The sludge and water were sprinkled on from a hand-held watering can. This was done on Munchong series soil (clayey) and continued from January to December, 1974. Ground conditions were observed, infiltration rates tested and soil analysis carried out.
Trial 2: A further trial was then commenced to test application of raw sludge at 5cm per month, applied twice per week to larger plots, 30m long by 4-4 m wide, between two rows of rubber plants. Such unreplicated plots were established on Munchong series soil (deep shale) in August, 1974 (under rubber clone GT 1) and on Malacca series soil (lateritic) in May, 1975 (under clone RRI M 623). The latter soil is considered poor, with low moisture-retaining capacity. These plots were increased in size to cover four interlines (about 0.135 ha of 50 trees) in January, 1976 and application rates stepped up to 10 cm per month. The plots were on undisturbed soil.
Similar observations were carried out as in Trial 1 and some effects
on the rubber were considered and compared with adjacent plots of similar size. A cover of dried solids soon developed on the treated plots, black and of a parchment-like consistency. It was the site of activity of numerous organisms. Insects and earthworms were common in the layer, fungus grew and the soil beneath was observed to be moist. Although no ground vegetation grew, due to continuous treatment, there was a noticeably larger amount of rubber root under the surface. There was rapid percolation of liquid at each new application and the solids soon dried (generally within a few hours), turned black and added to the layer. Once this happened deposits could not be reconstituted to resemble the original sludge. Rain percolated rapidly and applications in wet periods posed no problems of excessive run-off or failure of percolation of the liquid fraction.
The larger Trial 2 is continuing and at three years after commencement there is no evidence of excessive build-up of solids nor any change in the general appearance. The infiltration rates of water are shown in Table 2. They are inevitably subject to variation, from solid patchiness and changes in soil moisture, but it can be seen that there is no tendency for clogging of the soil, with fairly similar rates between treated and untreated plots in Trial 2. This would be expected from the greater activity of living organisms noted in the treated plots. The faster rates on treated plots in Trial 1 were surprising but were confirmed by several observations. These rates resembled those in a nearby plot with undisturbed soil, suggesting that removal of vegetation and top soil reduces the rates, whilst effluent goes some way to restoring them, rather than that the effluent increases natural capacity.
TABLE 2
COMPARATIVE WATER INFILTRATION RATES AT INTERVALS AFTER REGULAR APPLICATION OF SLUDGE (POMS) EFFLUENT FROM WATERING CANS. NEGEU SEMBILAN, 1975-6.
Centimetre infiltration in the first hour, at indicated period in months after the commencement of trial
Trial1 (Munchong series soil) 2 months 5 months 7 months 11 months 13months
Water 5 cma 11 30 34 29 27
Sludge 5cma 13 174 82 199 148
13 months at 12 months at 8 months at 17 months
5cm 5 cm at 5 cm and at 5 cm and
15 months 15 months
at 10 cm at 10 cm
Trial 2
Munchong
controlb 27 32
applicationb 64 21
Malacca
control 41 23
application 26 24
a5cm. 10cm, refer to rates of application; i.e. liquid equivalent to a 5 cm depth per month applied in portions twice per week.
b ‘Control’ plots, no liquid applied: ‘application’ plots, sludge applied
Manures and Sewage Sludges for Algal Growth
The ever increasing amount of wastes originating from human activities has introduced changes to the environment and in particular to the water quality. These wastes can be categorised into wastes from agricultural, domestic and industrial activities. Agricultural waste, including fertilisers, pesticides, pig and chicken manure, is the most important cause of water pollution in Hong Kong, especially in the New Territories. According to Bine pig manure and poultry droppings contribute 67.6% of pollutants entering water streams in the New Territories.
The estimated livestock population in Hong Kong at 31 March, 1978 was 493,360 pigs and 5,506,120 poultry. This great number of livestock produces waste equivalent to a human population of 1.45 million. Pig farmers wash their pig styes and let the effluent run into water streams without any form of treatment. Poultry waste is also thrown into water streams, as well as being used as fertiliser for land and fish ponds. Issac reported that in the northwestern New Territories of Hong Kong, pig and poultry wastes caused more than three-quarters of the organic pollution of the watercourses.
Animal wastes have been used for biogas production, nutrient enrichment for cultivating lower organisms, refeeding to livestock and land application. In the past, organised disposal of poultry and pig manures had been by application to land and fish ponds as fertilisers. Single-celled organisms growing in manured ponds served as supplementary food for fish and
resulted in large increases in the yield of fish per unit area of pond and sharp decreases in the weight of food which had to be added to produce a kilogramme of fish. Candida ingens, a pellicle-forming yeast, was found to grow on substrates derived from the anaerobic fermentation of monogastric animal (pig and rat) wastes. Later, the possibilities of recycling back through animals attracted some attention. Poultry manure was fed to sheep and poultry. Pig faeces was included in ruminant (cattle and sheep) diets. In Hong Kong, chironomid larvae are harvested from shallow water fields enriched with chicken manure for aquarium fish and carnivorous fish fry. Chicken manure has also been recycled as supplementary chicken feed and as fish-pond fertiliser. The treatment of swine waste by anaerobic digestion to recover methane
was covered in a comprehensive study by Boersma el al.
Sewage sludge is a liquid-solid mixture containing contaminants removed from wastewater by physical, biological and chemical treatments at sewage treatment plants. The wastewaters are treated either by aerobic activation or anaerobic digestion in order to inhibit the detrimental effect of sewage on the environment. With increasing knowledge of the sewage treatment process, the treated effluents will have a high water-quality but this will generate enormous quantities of sewage sludge. Its disposal gives rise to another major concern. Sewage sludge has been commonly applied on agricultural land. Recently, it has been noted that aqueous extracts of sewage sludge supported excellent growth of Chlorella pyrenoidosa and Chlorella salina. Ulva lactuca.
The properties of farmyard manure and sewage sludge have been studied by Yip. Farmyard manure contained a lower level of various heavy metals but a similar content of essential nutrients and would be safer for reutilisation than sewage sludge, especially if the products were used for animal feed.
Recycling of these agricultural wastes and sewage sludges not only reutilises the rich nutrients (e.g. nitrogen and phosphorus) in the wastes but also minimises the problem of water pollution in the environment.
The use of algae to convert waste materials to a usable feed grade protein supplement has received the most attention and appears to have the most promising future. Chlorella has been studied largely because it can be synchronised easily with a light-dark cycle. Interest has been renewed in producing single-celled protein by mass culture of unicellular algae. Algal systems can be used for both pollution control and protein production by directly recycling the nutrients in wastewaters into biomass.
The purpose of the experiment described in this paper was to compare the suitability of four wastes—activated sludge, digested sludge, chicken manure and pig manure—for the cultivation of Chlorella pyrenoidosa, a unicellular green alga.
Tower Digestion of Pig Waste
Currently marketed farm-waste digesters usually consist of cylindrical tanks made of concrete, ferrocement or mild steel. They are heated electrically or by burning digester gas, and mixed by small, high-speed propellers, large low-speed angled blades, or recirculated digester gas, with or without a draught tube. In spite of this diversity in construction detail, from a process engineering point of view all are essentially completely mixed reaction vessels with semicontinuous feed and effluent flows. As there is no provision for recycle of active biomass or solids undergoing digestion, all approximate to the chemostat of biochemical reactor theory. Process development by equipment suppliers appears to centre on the prevention of accumulations of floating or settled solids which can cause operating problems such as blockage of pipes, loss of effective digester capacity and poor contacting of fresh waste with active
biomass.
Waste retention times ranging from 10 to 20 days for heated digesters to 20 to 50 days for unheated units are required for acceptable digestion efficiency. If waste hydraulic retention time (HRT) is reduced below 10 to 15 days, gas yield falls off, and below 5 to 7 days digestion becomes unbalanced and methane production and waste stabilisation cease. The reasons for this reduction in digestion efficiency are the slow rates of hydrolysis of lignocellulosic fibre particles and the classic increase in unreacted soluble waste level as waste retention times approach the minimum doubling times of key bacterial groups, the acetoclastic methanogens and the obligate hydrogen-producing acetogens. These doubling times have been measured as 1-2 days and 3-6 days respectively under optimal conditions in pure culture, and are likely to be longer in pig manure digesters with inhibitory levels of ammonia nitrogen. Attempts to operate completely mixed digesters at HRT below these doubling times invariably lead to the accumulation of fatty acids and process failure.
In order to reduce waste HRT (and therefore digester size and cost) without sacrificing digestion stability and the efficiency of waste breakdown, it is necessary to retain active digester bacteria and fibrous waste solids for periods longer than the HRT. This requires a means of solid-liquid separation within the digester. For soluble wastewaters this has been achieved by gravity settling of bacterial flocs (contact and upflow sludge blanket processes), floc entrapment (upflow filter) and by bacterial attachment to fixed or mobile supports (stationary fixed film and expanded and fluidised bed processes).
While attached film processes have enabled stable digestion of pig and dairy manure at HRT as short as 48 h and 3 h this has been at low solids content and accompanied by a reduction in gas yield, the result of short fibre retention times. And, in order to apply floc-based processes to manure digestion, fibre has first been removed by screening or in a separate hydrolysis/acidification digester. This has been at the expense of gas yield or added process complexity.
Fig. 1. Industrial tower fermenter for ethanol production.
(Redrawn from Greenshields & Smith, 1971).
The tower fermenter is an upflow microbial floc reactor similar in-many respects to the upflow anaerobic sludge blanket process. It had previously been used in fermentations in which short hydraulic retention times and one microbial retention times were maintained with flocculent yeast or filamentous fungi. Feed enters the lower base, passes upwards through a dense fluidized region of microbial flocs or pellets and leaves via a settling region where entrained biomass separates and returns to the reaction zone. Since tower fermenters had previously only been used with soluble feeds, in the present work provision had to be made for the removal of undigested solids. In addition, as significant natural bacterial floc formation in pig manure digesters had not previously been reported, it was considered necessary to create bacterial flocs by artificial means.
This paper describes the search for an effective flocculating agent, the design of a tower suitable for manure digestion and the performance of this tower in terms of the accumulation of digester bacteria and of waste fibre, and in terms of gas yield and digestion stability.
Chemical Composition of Palm Oil Mill Effluent
Palm oil mill effluent (POME) is a voluminous, high BOD, liquid waste. About 2.5 tonnes are produced for every tonne of oil extracted in an oil mill. The effluent comes mainly from the sterilisation and clarification sections of the milling process. POME is essentially a colloidal dispersion of biological origin. This property, coupled with its high BOD loading and low pH, makes POME not only highly polluting but also extremely difficult to treat by conventional methods. Consequently, direct discharge of effluent into receiving waterways poses a serious threat to the environment.
Chemical analyses of POME with respect to its proximate composition have been carried out previously. Values for ash, crude protein, ether extract, crude fibre and calculated residues, called ‘nitrogen-free extract’, were reported. Such analyses are usually carried out for feedstuffs in relation to their utilisation as nutrients. Information on the mineral contents of POME is also available. However, all these data have been collected mainly from the mixed effluent (that is, POME), less is known of the chemical composition of the individual discharges from the different sections of the milling process, although some results on the clarification sludge and steriliser condensate with respect to mineral content and carbohydrate constituents were reported by Hwang et al.
A thorough chemical characterisation of POME is important not only to provide vital information on its pollution potential but also to understand its properties in relation to formulating waste-utilisation programmes and efficient waste treatment processes. This is especially so in view of the increasing emphasis placed on waste utilisation for energy-production purposes. A case in point is the production of biogas or liquid fuel by fermentation processes. Of no less importance is the possibility of recovering from the waste, or converting it into, useful substitutes for animal feed. All these processes require a detailed knowledge of the chemical constituents of POME which unfortunately is not readily available in the literature. Furthermore, there seems to be no data published on the partitions of the various chemical constituents between the soluble and the particulate fractions of POME. Earlier work initiated in these laboratories has indicated that it is possible to fractionate POME. by high-speed centrifugation and thus the distribution of chemical constituents between the fractions can be delineated.
METHODS
Sampling
Effluent samples were collected from a 30 ton/h mill. Raw POME was sampled from the discharge point after the oil trap (the pit). The clarification sludge samples were collected immediately after the battery of centrifuges in the clarification station. All samples were transported to the laboratory and analysed on the same day.
Fractionation procedure
A slightly different procedure from the one described previously was employed for the fractionation of the effluent samples. Samples were first centrifuged at 15 000 g at 5 °C for 30 min. This removed the water-soluble constituents from free oil droplets and suspended solids in the effluent. The sediment was redispersed in twice-distilled water and recentrifuged. The washing procedure was repeated three times. Tlie washings were pooled together with the original aqueous supernatant and concentrated to a small volume before use. Part of this was dried in an oven at 105 °C to give the dissolved (soluble) solids content. The washed sediment was also dried at
105 °C. This gave the particulate fraction and the weight was the suspended solids
content.
Assays
Total solids content was determined by oven-drying of the whole sample at 105°C. Ash contents were determined by ignition in a muffle furnace at 525 °C according to the AOAC methods. The oil content was determined on dried samples (original and fractionated) by extraction with hexane for 6 h followed by drying the oil to constant weight. The extracted sample was then further extracted with ethanol/benzene (1:2 v/v) according to the TAPPI method. The sample thus treated was termed extractive-free sample and used for further analyses to be described later.
Free glucose and reducing sugars were determined by the glucoseoxidase-peroxidase method of Worthington and the Shaffer-Somogyi micro-method of the AOAC, respectively. Glucose was employed as a standard for both determinations. Total glucose, total reducing sugars and total carbohydrates were estimated after acid-hydrolysis of the dried samples. Typically 0.35 g of sample was digested with 3 ml of 72 % sulphuric acid at 30 °C for 60 min followed by diluting with water to 84 ml and autoclaving at 121°C for another 60 min. The resulting mixture was centrifuged and the clear supernatant analysed for glucose and reducing sugars by the methods mentioned above and for total carbohydrate by the phenol-sulphuric acid method of Dubois et al. again with glucose being used as a standard,
Pectin was estimated by precipitation with alcohol as outlined by McCready. Extractive-free samples were used for acid-insoluble lignin and holocellulose determinations by the TAPPI method and the method of Wise respectively. Phenolics were also estimated spectrophotometrically on the dried extractive-free samples by the method outlined by Vered using pure Klason lignin as standard, α-, β- and γ-celluloses were determined by the TAPPI method. The determination of pentosans was carried out using the TAPPI method. The analyses of starch and Kjeldahl nitrogen were according to the AOAC methods.
Each reported analysis was the average of four determinations from samples collected on different occasions.
Methane from Cattle Waste
Disposal of liquid manure by flush systems can efficiently transport animal manures with initially up to 15% solids. Labor requirements are reduced in such systems, although this method can result in a large volume of dilute waste. Solids separation of liquid manure by particle size or density can remove up to 30% solids. The remaining dilute liquid and fines fraction is then available for recycling as a waste diluent, irrigation and fertilizer resource, or substrate for additional processing.
Methane production from livestock sources has been shown to be an easily established fermentation process and one-third of the total energy content is released in the form of methane.
Methane fermentation of fractionated dairy waste has been limited, due in part to the low fermentable-solids content and relatively high capital construction costs for the fermenter system. Rorick et al. (1980) demonstrated similar methane production from both anaerobically fermented, unfractionated and separated, liquid-effluent dairy waste. Similarly, demonstrated that liquid solid separation has little effect on methane production from dairy waste per unit of volatile solids (VS) added under specific conditions.
One low-cost approach to methane fermentation of low-solids dairy waste is via a tubular fermenter. Advantages of a tubular fermenter. Which consists principally of a pipe and pump, have been reported. The highly developed technology and ready availability of construction materials warrant additional consideration of this fermenter for use with dairy waste and other low-solids content wastes.
The objective of this study was to evaluate the methane-fermentation characteristics of fractionated dairy waste in a tubular, single-pass fermenter.
METHODS
Herd characteristics
Lactating Holstein-Friesian and Jersey cows were confined outside on solid concrete floors in a free-stall, partially covered lot. The cows (65 to 103 in number) were maintained in the lot and fed, twice daily, a dairy ration containing 50% corn silage and 50% concentrate (corn, cottonseed meal, soybean meal, citrus pulp, stillers grains and mineral supplement). Water was provided ad libitum.
Fermenter substrate
Wash water from the milking parlor was used as a waste diluent in a gravity-flush waste disposal system. The cow waste was diluted, flushed during milking periods, and collected in an underground storage tank. The diluted waste (<12h old) was then agitated and passed over an inclined gravity separating screen (Bova-sieve, Agpro Inc., Paris, TX). The separated solids from the screen were collected and used as bedding material in the free-stall area or disposed of on the land. The liquid and fines fraction (LF; material passing through the inclined screen with 1.0 x 6.0 mm sieve size) were transferred and stored (<3-0 h) in a continuously agitated mixing tank. Sufficient LF substrate for three fermenters was collected daily in the mixing tank prior to simultaneously metering into each fermenter. The fermenters were fed LF within 3 h daily and an equal volume of effluent was removed by displacement.
Fermenters
Three pilot-scale (453 liter), horizontal (< 3% slope), unagitated, tubular fermenters were constructed from 30.5 cm diameter x 609.6 cm steelpipes. Machined end-plates were fitted with substrate feed and effluent ports; sample and gas ports were installed at each end. The inlet feed port was located in the center of the fermenter inlet end-plate. The effluent port was located at the top of the fermenter outlet end-plate to retain settled solids. Additional sample ports were installed at 0.5 m distances along the entire length of each fermenter. These ports allowed sample collection from the center of each fermenter. Fermenter temperature was maintained with a recirculating, thermostatically controlled (<2.5°C), solar-heated hot water system. The heat-exchanger consisted of a copper tube (0.95 cm diameter) coiled (1 coil/10 cm of fermenter length) around the exterior of each fermenter. Mounting the heat exchanger coil on the exterior of the fermenter eliminated coil coating which can reduce thermal exchange performance, and fermenter capacity was not reduced due to coil displacement. The exterior was covered with 10 cm of polyurethane foam insulation and sealed with an epoxy paint. Typically, the average time required to reach fermenter operating temperature at 40°C after feeding was 3.8, 3.5,.3.2,0.0 and 0.0 h at HRTs of 1, 2,4,8 and 10 days, respectively.
Experimental
The fermenters were initially filled with water and flushed with nitrogen (5 liters/min) for 30 min. Temperature was established for each fermenter and maintained at 20, 30 or 40°C throughout the experiment. Single measured daily feedings, equivalent to hydraulic retention times (HRTs) of 1,2,4,6,8 and 10 days, were performed for each temperature condition. The fermentations were initiated at a 10 day HRT, and uniformly decreased to the shortest HRT. Microbial seeding from an operating fermenter was not carried out. Substrate feed, effluent and gas samples were collected daily and analyzed as described below. Fermenter ports were sampled at steady-state (minimum of 6 times the HRT). The sampling was performed 1 day following the last feeding by first withdrawing and discarding 2 liters of substrate and immediately collecting-0-5 liters of substrate for analysis.
Subsamples (approximately 100 ml) were refrigerated and analyzed in duplicate (after <7 days) for total solids (TS), volatile solids (VS) and fixed solids (FS) by AOAC methods. Selected samples were analyzed in triplicate (after <48h) for chemical oxygen demand (COD) by Standard Methods. Additional subsamples (approximately 15 ml) were collected in glass sample bottles and stored at — 18 °C until analyzed. The supernatant obtained after centrifugation (15 000 x g for 15 min at 6°C) of the preserved subsample was used for determining the concentrations of volatile acids (VFA) by gas chromatography. The VFA were determined in duplicate using carbowax, 20M/ 0.5%H2PO4 on 60/80 carbopack B (Supelco Inc., Bellefonte, PA) by the procedure of DiCorcia. Gas production was continually measured with a wettest meter and volumetrically corrected to standard temperature and pressure daily prior to substrate addition. Gas samples were analyzed for methane and carbon dioxide content by gas chromatography. Duplicate samples (1.0 ml) were injected onto car-bosieve S 120/140 mesh (Supelco) and separated isothermally at 121ºC. The VFA and gas samples were compared to similarly treated and processed standards.