Small-scale processors, when well-dispersed, will have minimal impact on the environment due to the low discharge rates. Large-scale processors are more easily monitored, as they are few in number, and will have the financial and technical resources to deal adequately with waste management. Environmental pollution from medium-scale cassava processors is the more difficult to deal with. At this level of development, plant concentration and individual plant discharges can easily combine to create significant environmental impact. Yet, each processor has limited resources (financial and technical) to deal with environmental waste. Compounding the problem is our limited ability to develop a targeted, low-cost treatment processes for small- and medium-scale cassava processors given our present knowledge. Basic knowledge on environmental impact is still lacking, and without this information it is difficult to develop appropriate technologies that do not negatively impact the environment.
Given that suitable technologies are available or can be developed, the problems of pollution from cassava processing are more social and economic in nature than technological. Interventions, usually from the government, are required. Most governments recognize the need to control waste produced by cassava factories, but they are equally aware of the economic risk involved in such a strategy. Accessible technologies for most scales of processing are available; however, the cost of implementing the technology is, in many cases, prohibitive. According to starch processors, the installation of pollution control devices can be 20-50% of the total investment cost of a large-scale factory. Full implementation of strict environmental controls too quickly can have negative consequences, forcing the industry to forfeit its competitiveness. In many countries two policies operate - those applied to established processors and stricter requirements imposed on new entrants. The scale of a processing enterprise is important; the smaller number of large-scale processors is, logistically, easier to regulate than a large number of small processors.
In Brazil, the government has intervened through legislation and recommendations for waste treatment technology. The legislation passed in 1992 requires all processors to reduce waste water BOD by 80%, or render it such that the BOD of the water course, after receiving the treated waste, is less than 5 mg/l. Unfortunately, the government seems to be slow in enforcing the regulations (G. Chuzel, personal communication). In contrast, the government of Thailand enforces 'clean-up standards' that must be followed by all factories. The government of Colombia, through a regional agency (CVC) is now conscious of the need to develop a cleaner technology for sour starch extraction. Joint projects are currently being carried out to develop appropriate technologies for waste water treatment, such as bamboo filters and lagoons. However, since the implementation of this stricter legal framework for waste management in 1999, the livelihood of small-scale processors is being threatened as non-complying units have to close.
Dealing with environmental problems resulting from processing is generally regarded as a necessary expense with no direct return. This does not have to be the case. New opportunities can be realized through tackling of environmental problems, especially when the industry is still developing. At that time, strategies do not have to be dovetailed into the restraints of an existing structure, but can be built into the development process.
A number of approaches are used to 'clean' environmental problems created by cassava processing. For a processor to use such 'end of the pipe' technology, developmental work is required. The investment in new, or modification of existing, technologies would be more efficiently directed to investing in improving, or at least preventing the problem. If the only choice is to deal with an environmental problem, it is recommended that efficiency be compromised and the most suitable available and fully developed technology chosen.
In cassava-processing areas dominated by small to medium-sized enterprises, none of the producers has an effluent treatment system capable of treating waste to permissible levels. Most waste is discharged without any treatment. Some factories store the effluent in settling tanks for a number of days before discharge; in this case, BOD and COD levels are reduced and most cyanide evaporates. Others dig effluent pits and channels, which, depending on the conditions of the underlying soil, can provide an adequate means of disposing of waste.
Technologies available for treating cassava processing waste (from small and large factories) include:
Land filling. This is sometimes used to dispose of solid residue.
Use as animal feed. Use of solid waste for further processing into animal feed is practiced in many regions, including South America (Pereira, 1987), Africa (Montilla, 1977; Adebowale, 1981, 1985; Tewe, 1987; Adeyanju and Pido, 1978; Obioha et al., 1985) and Asia (Hutagalung, 1983; Manilal et al, 1991). This is widely practiced, but does have limitations depending on the availability of other resources. For example, in Cong Hoa village, Vietnam, use of waste as animal feed is restricted by the amount of available protein to supplement the carbohydrate, and by a lack of knowledge. One novel approach has been to develop an integrated cassava starch processing/bioreactor/aquatic plants/animal production/system. In this system, the cassava is indirectly used as a source of animal feed (Viet, 1998).
Ensiling of solid residue. Ensiling lowers the cyanide level to one that is non-toxic, leads to a reduction in pH to 4.0 and allows lactic acid to build-up. The product is subsequently used as animal feed. (Nguyen Thi Loc et al., 1997).
Fermentation of cassava peel. Cassava peel is a major residue in some countries. Utilization of the peel is limited by its low digestibility and toxicity from extremely high levels of hydrocyanic acid. Fermentation not only reduces toxicity, but the enzyme-resistant ligno-cellulose material is converted into a more digestible substrate. Following fermentation, cassava peel can be formulated into pig and poultry feed (Ofuya and Obilor, 1993).
Use of waste water for irrigation. A traditional method of dispersing of processing wastes is to return it to the land as irrigation water. This requires careful monitoring to ensure that long-term soil degradation does not occur. Use of waste water for irrigation or as a source of fertilizer may be restricted as the high HCN content can have a negative effect on plant growth (Taesopapong and Bhanuprabha, 1987; Bengtsson and Triet, 1994). In one study cassava waste water was used as a direct fertilizer for duckweed at a dilution rate greater than 60% (waste in water). All duckweed died; a dilution of 10-20% was required for plant survival.
Infiltration of waste water into the soil. In many sites, especially those where small- and medium-scale processors predominate, waste water is minimally treated by channeling into shallow seepage areas, ideally situated away from natural water courses and groundwater abstraction points.
Storage in aerobic or anaerobic lagoons. Some of the starch factories in Brazil, India (in Kerala), Vietnam and Thailand have built anaerobic and aerobic lagoons to treat waste water before disposal. These units are of varying efficiency, require a large area of land and are capital intensive. In anaerobic digestion of cassava waste, cyanide is released in the fermentation liquor and then liberated by enzymatic and non-enzymatic reactions. The removal of cyanide has been shown to be sufficiently fast to maintain a cyanide concentration in the reactor, which is non-inhibitory for methanagenic bacteria (Cuzin and Labat, 1992).
Anaerobic digesters. Traditionally anaerobic digestors have been used for the treatment of agricultural wastes. These processes require large tanks or bioreactors and long retention times of 20-25 days. Recent advances in treatment technology and knowledge of microbial process control have led to the development of high-rate anaerobic treatment processes, some of which are being contemplated or used by the cassava starch industry. High-rate anaerobic treatments make use of microbial films to achieve high cell residence time. These processes operate in low hydraulic retention time and can process large amounts of organic material. Biofilm processes used by the industry comprise different engineered configurations, such as fixed bed, moving bed, fluidized bed, recycled bed and upflow anaerobic sludge blanket (UASB). All these reactors can handle loads up to 20-30 kg COD/m3/day). All of these processes require a relatively small reactor size, and a vastly reduced requirement for land and capital. The following systems are frequently used:
a. Fixed bed: (anaerobic filter, packed bed filter, packed bed, submerged filter, stationary fixed film reactors). The principle of operation is that the support material is also the surface for attachment of the microorganisms and can act as an entrapment mechanism for unattached flocs. Many support types are used, including quartz, plastic, clay, oyster shells, stones, polymer foam, activated carbon and sand.
b. Fluidized bed reactor: In this type of reactor most of the biomass is attached as films to small-sized inert media. The biomass-covered particles are lifted (fluidized) by the high vertical velocity of the incoming waste. Various support materials are used, such as sand, PVC, and granular activated carbon.
c. Upflow anaerobic sludge blanket (UASB): This type of reactor consists of a dense bed of granular sludge (microorganisms) placed in a reactor that is designed to allow upward movement of liquid waste. Waste water entering at the reactor bottom is distributed across the cross-section and flows upward through the bed of sludge granules retained in the system. Sufficient upflow velocities are maintained in the reactor to facilitate sludge blanket formation and to provide a greater surface area for contact between sludge granules and waste water.
Alazard (1996) describes a laboratory-scale evaluation of different anaerobic treatments for waste-water from starch extraction in Colombia. These studies include comparison of UASB, UASB with phase separation, a "transfilter" process (Farinet, 1993) and a rustic biofilter "bamboo horizontal flow". All these processes seem to be efficient in lowering the organic load of the waste. In all cases, efficient biodegradation (up to 90%) at an organic loading of 5g COD/l/day is attainable, while maintaining a hydraulic retention time of nine hours. The production of biogas with 70% methane is about 350 l/kg of COD removed. Given the low investment costs of the bamboo horizontal flow method, and its comparable efficiency with other more expensive procedures, this technique has potential for small-scale processors.
In Brazil, two units of anaerobic treatment with phase separation have been installed in farinha factories, each with a capacity of 10 to 20 m3 per day (Cereda, 1996). Despite encouraging results in terms of removal of organic material (reduction of 80%), the high cost of maintaining these units has resulted in their decommissioning.
The biological treatment of waste water is based on a simple process, in which mixed populations of microrganisms utilize the nutrients in the waste. Their efficiency is extended by use of chemical engineering techniques that allow the basic process to be intensified and accelerated. This gives the range of biological treatment systems currently in use for treating agricultural waste water (Degremont, 1991).
Waste water containing both organic material and a source of nitrogen is brought into contact with a dense population of microorganisms. Sufficiently long contact times are engineered into the process so that the microorganisms can break down and remove the pollutants (organic material) to a desirable level. This process can occur in the presence of oxygen (aerobic) or absence of oxygen (anaerobic). In addition to clean water, the nature of the other products will depend on the process. Anaerobic systems produce biogas (methane, carbon dioxide, and small amounts of hydrogen sulfide and ammonia) and biomass (microorganisms); aerobic systems create carbon dioxide and a large amount of biomass (microorganisms).
Lagoons are the simplest form of biological treatment, and the type of lagoon used (anaerobic, facultative, aerobic) is dependent on the available area and amount of waste to be treated. Because of the high organic content of cassava processing waste, the first lagoon is usually anaerobic.
Cost estimates of lagoon systems need to take into account the land area required and the soil type. A lagoon or ponding system is cheap to construct but requires a large land area. If the lagoons are constructed in permeable soil, the need for lining, consisting of either clay or synthetic material, will add significantly to construction costs.
Lagoon systems are normally operated at low rates with organic loading ranging from 0.2-0.35 kg BOD/m3/day. Because of the size and configuration of the lagoons, they are quite difficult to control and monitor. Energy required to operate a lagoon system is minimal. Electricity is only required to run the pumps; gravity flow is exploited where possible. Different types of lagoons include:
Aerated lagoons have an aerobic surface layer, generally maintained by floating surface aerators, with an anaerobic zone below. A large amount of organic material can be applied; thus, maximum organic reduction with minimal odor and associated nuisances can be achieved.
Activated sludge is an anaerobic process, named because the treatment takes place in an activated (containing microorganisms) and intimately mixed liquor. This increases the mass of organisms available for waste reduction.
Anaerobic digestion is a complex two-stage biological process in which the organic matter is reduced in an anaerobic environment. Anaerobic digestion facilities may range from a simple, unmixed, unheated open tank (low rate digestion) to a mixed and heated covered tank (reactor), incorporating collection and utilization of the gas produced, followed by a secondary digester for liquid/solids separation (high rate digestion). An anaerobic lagoon is essentially a crude uncontrolled anaerobic digester.
The lagoon system is the most popular treatment system used by cassava starch processors for treatment of waste water. Popular in Thailand are "no-discharge" systems that consist of up to 20 ponds. The combined waste water is often collected in a storage pond from where it is pumped through a screen into a pretreatment pond. After this, the waste water is pumped through a series of treatment ponds, the first 2-3 being anaerobic, where organic substances are successively degraded by natural breakdown processes. The amount of organic material that can be added is about 800-1,000 kg COD/ha/day. Thus, in order to treat 6,000 m3 of waste water per day (a typical discharge from a Thai cassava factory) with a COD concentration of 14,000 mg/l, requires a land area of about 100 ha. This will involve 20 lagoons, each with an area of 5 ha. Typically, residence times of 350-400 days are necessary. An alternative system will discharge the water after a period of 100-200 days. This system differs in that during later stages the ponds are aerobic, to accelerate the breakdown of organic matter. The treated waste water will have a BOD as low as 15 mg/l and can be used for irrigation (see Appendix 3). This system requires fewer ponds (10) and less land area (8 ha).
These technologies for treating the waste of cassava processing, whilst removing the environmental problem, incur investment costs from which there will be no direct return. Much research has been conducted to address this problem.
Transforming the waste offers the possibility of creating marketable value-added products. This is successfully exploited by the maize wet milling industry, which generates about 40% of the total revenue from the by-products of maize starch processing. It is a powerful incentive for cleaning waste streams. This strategy can relieve some of the financial burden incurred by waste treatment. The composition of the raw material is important, and for cassava limited opportunities exist. There should also be sufficient market opportunity for the product. In most processing zones, cassava waste is treated as a low value product, but is sold.
In Thailand cassava peel is utilized as a medium for mushroom cultivation or is used to produce compost. In Guangxi, China, pulp, which is generally used as an animal feed, is used as a raw material for the production of ethanol (Henry and Howeler, 1996). In Vietnam, solid waste (mainly pulp) is sun dried and used as fuel for production of maltose in the same village. In Thailand, most fibre waste is sun dried, mixed with ground chips and pressed into pellets for export to Europe. As starch production increases and pellets exports decrease this may not take care of the problem in the future.
The literature is replete with 'novel' technologies for treating agricultural waste, including that produced by cassava processing, many involving fermentation by bacteria or yeast for production of a biofuel, such as ethanol (Tanticharoen et al., 1986; Abraham and Muraleedhara, 1996) or biogas (Tanticharoen et al., 1986). Other technologies involve the production of single cell protein from cassava waste (Manilal et al., 1991) or spirulina (Tanticharoen et al., 1991). Adoption of such technologies should be with caution. Markets are often poorly identified for the ultimate product and development work (scaling up technology from laboratory to field) must be financed.
A phased approach to the introduction of an effluent treatment system is recommended. Rapid progression from no treatment to full treatment may be unrealistic, both for technical and financial reasons. The staged approach allows time for the operator to adjust to new equipment and solve new problems, increasing the overall project success.
Preventing the waste problem from occurring in the first place, is the ultimate solution, and one that is available for some cassava processors.
The most appropriate form of prevention for the starch processing industry is to reduce the water requirements and hence the amount discharged. This can be achieved simply through more efficient use of water through re-cycling. Some factories use water from the sedimentation tanks for root washing (Figure 18). The lower pH of this water also has the advantage that it better removes the soil and debris from the roots.
ERM (1996) has conducted field studies of small- to medium-scale starch and sago processors in Salem district of India to investigate the potential application of hydrocyclone technology for water conservation. The use of this technology led to a reduction in the water requirement by 50-60%, and a reduction in waste water volume by 40-50% (Marder, 1994). The pollution load of the waste water was also reduced by 50% (Trim and Marder, 1995). This still needs to be used together with an 'end-of-pipe' waste water treatment solution. However, the amount of water to be treated is markedly reduced, and thus the demand on the treatment technology is lower.
Water can also be suitably treated with compounds such as alum (concentration of 80 mg/l water) to precipitate the suspended solids and reduce BOD; afterwards, the water can be reused in a recycling system (Ong and Loh, 1986).
While cassava processing is generally perceived as polluting the environment, very little research has been conducted to quantify the levels of pollution, and determine the magnitude and significance of this pollution on people and the environment. Also, little research has been conducted to develop efficient and cost-effective ways to reduce pollution, especially by small- and medium-size processors. Specific gaps in our knowledge exist on:
1. Criteria to identify pollution "hot spots", and a model to determine the maximum number of processors able to operate in a particular geographical area before water availability and/or pollution becomes a major problem.
2. Relationship between soil characteristics and the effect of processing on groundwater supply and quality.
3. Effect of key cassava processing unit operations, such as retting, squeezing or grating, on the environment.
4. Relationships between processing parameters and the resulting COD, BOD, HCN and SO2 concentrations of various waste products.
5. Cost-effective ways of reducing COD and HCN in the waste water of small-scale processors.
6. Value-adding of waste products, such as fermentation and protein enrichment of cassava pulp to be used as animal feed, especially for small-scale processors.
From the above discussion the following conclusions and recommendations can be made:
1. With the exception of starch extraction, most other cassava processing does not require large quantities of water. Cassava starch processing should be located in areas of adequate water supply. But even in areas with a limited water supply, cassava starch processing normally does not seriously deplete natural water resources.
2. There are few studies investigating the effect of cassava processing contaminating the ground-water supply. From those undertaken, evidence is not conclusive. Contamination is most likely to occur in regions where a large number of processors are concentrated in a small geographical area. These potential pollution "hotspots" should be identified, and groundwater quality, along with complaints about public health problems, should be closely monitored.
3. Water consumption should be minimized by the use of hydrocyclones or other water recycling systems. These systems also increase extraction efficiency and reduce the amount of waste water produced.
4. Waste water should be either contained within the premises of the factory or treated before release to water sources outside. Depending on the size of each processing operation and their spatial distribution, the following waste water treatments are recommended:
a. For small-scale and isolated processing units:
Seepage pits are the cheapest solution, guaranteeing a limited degree of effluent reduction. Such pits can provide a viable and easily replicable means of containing wastes. Soil conditions should be considered carefully before choosing a site for the pits.
b. For small-scale but densely clustered processing units:
Common effluent treatment, which may include storage in aerobic or anaerobic lagoons before release to outside water sources.
d. For medium- to large-scale starch factories:
Individual effluent treatment, such as storage in aerobic or anaerobic lagoons, either alone or in combination with various types of anaerobic digesters. After treatment the water can be used for irrigating crops or safely released to rivers or streams.
5. Solid waste such as cassava peels, fibrous residue and starch residue can be disposed of as follows:
a. Land filling.
b. Value-adding by drying or ensiling of peel and fibrous residue to produce animal feed (to be mixed with a protein source).
c. Production of alcohol or compost.
6. HCN released from cassava roots during processing will be present in:
a. Cassava peel. These should be ensiled or fermented to reduce toxic levels of HCN before use as animal feed.
b. Press water (gari and farinha production). High concentrations of HCN in press water can be toxic for humans, plants and animals. The press water must be stored or fermented to reduce the HCN content by evaporation before release or further utilization.
c. Waste water. By storing waste water in seepage pits or lagoons, most HCN will evaporate and be rendered innocuous.
d. Vapor and water sprays during processing. Work areas where cassava roots are grated in high-speed graters, or where cassava products are toasted or boiled, must be well-ventilated to prevent high concentrations of HCN in the air, affecting the health of workers.