Cassava as an adhesive's ingredient in Ghana

Introduction

This section provides a more detailed analysis of one specific market for cassava starch -that of the adhesives market in Ghana. As noted above, the paperboard and plywood industries of Ghana are significant users of starch, flour and starch-based products. In 1996 the paperboard industry used 420 tonnes of starch-based adhesives in the manufacture of corrugated board. In the same period the plywood industry used 1,134 tonnes of starch and 1,200 tonnes of food grade wheat flour as extenders for synthetic wood glues. The paperboard and plywood industries account for 37% (excluding wheat flour) of the market for starch and starch-based products in Ghana (Graffham et al. 1997).

There are currently six manufacturers of starch-based adhesives for paperboard in Ghana, half of which prepare some or all of their products using locally produced cassava starch or flour. The paperboard sector comprises four companies, which use starch-based adhesives (SBA) in the manufacture of corrugated board. Imported adhesives based on maize starch account for 55% of the market and cost between US$900 and US$1,260/tonne depending on the country of origin. The remaining 45% of the market is accounted for by a local manufacturer of cassava-based adhesives, who sells at the equivalent of US$430/tonne (Day et al. 1996). The imported materials are prepared from pure maize starch. The local product instead, is prepared from rather poor quality cassava flour derived from unpeeled cassava roots that are sun dried, coarsely milled, screened, and blended with other adhesive components. Both local and imported products are supplied as ready mixed dry powders in 50 kg bags.

Three out of the four-paperboard factories use local as well as imported adhesives. The local product is favoured because of its low price and ready availability. However, all users commented unfavourably on the quality of the local product. Locally produced SBAs form weaker bonds, have a short shelf life, contain too many contaminants, and are not finely milled. Users overcome bonding problems by blending local and imported products together (Day et al. 1996).

The plywood industry in Ghana comprises eight large-scale factories, which use imported synthetic resin-based glues in the manufacture of plywood sheets. These glues cost US$2,220/tonne. To reduce costs, synthetic glues are mixed with an extender that can be either imported maize starch (US$650/tonne) or food grade wheat flour (US$500/tonne). Typically 50 kg of synthetic glue will make 55-60 1/8" plywood sheets, with an extender this increases to 80-85 sheets of 1/8" plywood sheet. For each 50 kg batch (of synthetic glue), either l0 kg of maize starch or 25 kg of wheat flour is required (ibid). Several factories have tried locally produced cassava flour and starch as extenders but found these to be of poor quality and discontinued use. Locally produced flour was not milled properly, insufficiently dried and contained many insoluble impurities that caused blistering in the plywood sheets. One manufacturer claims that local flour caused his percentage of rejects to rise from 1 to 7%.

General specifications for SBAs and plywood glue extenders are given in Table 25. Users of these products are not generally much concerned with cyanide content, microbiology, colour, taste or odour but would not expect a product to show signs of mould. Users of paperboard adhesives may have specifications for shelf life and bonding characteristics but are normally flexible if the price is attractive.

The Research Institutes' approach

The Natural Resources Institute (NRI) and Food Research Institute (FRI-Ghana) have followed two approaches for production of cassava chips (Figure 5) that can ultimately be milled into flour for industrial use at a centralized processing facility.

The first approach, commonly referred to as the farm-based option, focuses on field-based production of cassava chips by farmers; the second makes use of a centralized processing facility with higher levels of mechanization and mechanical drying. The NRI/FRI process uses modified IITA mini chippers as these provide chips that dry in the shortest possible time. The farm-based option has the advantage of being low cost (US$65/tonne)^[5] and produces chips of sufficient quality, if some measure of quality control is in place. The centralized option for chip production is more suited for large-scale production of consistent quality chips, albeit at much higher cost (US$217). In many cases field-based production of chips would be appropriate followed by milling and formulation (for SBAs) in a centralized facility such as a flourmill or glue factory.

Starch-based adhesives used in the manufacture of paperboard consist mainly of starch flour blended with certain chemicals (Table 26). The essential ingredients in starch-based adhesives (SBA) are starch and flour, gelatinisation modifier (sodium hydroxide), viscosity enhancer and stabiliser (borax) and preservative (sodium formaldehyde). A basic formulation for 1 tonne of SBA in dry form could be as follows:

Using the prices given in (Table 26), this product would cost between US$204 and US$350 per tonne depending on the cost of cassava flour. The amount of borax and sodium hydroxide must be determined experimentally so as to provide a starch-based adhesive (SBA) with the correct viscosity and pasting temperature to meet the user's specification. As the range for these ingredients is quite wide (Table 3.7), costs could be considerably higher in practice.

In addition to the basic ingredients, fillers such as clay or coconut shell are often added, and in some plasticisers, lubricants, bleaching agents and antifoams may be included.

All of these will greatly increase the cost of the product and should only be used if essential to meet the user's requirements.

The chemical cost quoted in Table 26 only applies to relatively small quantities purchased for trial purposes by NRI/FRI. In commercial practice, prices would be much lower because chemicals are normally sold in 16-20 tonne containers at negotiable rates. Price is normally determined by quantity, customer reliability, prevailing exchange rate and shipping cost (Graffham and Dziedzoave 1998).

This market opportunity has been mainly influenced by changes in national policy and economic factors. But it would also be true to say that improvements in the network of major roads linking inland areas of Ghana with the coastal cities and major ports has made cassava more attractive as a raw material for locally produced starch based adhesives and glue extenders.

Current production of cassava based adhesives centres around Central and Ashanti Regions, which have large supplies of cassava, good road infrastructure and easy access to users of these products in Accra, Kumasi and Tema. Cassava chips are sourced directly from the farm and milled and formulated at the adhesive factory. Adhesive suppliers produce and supply adhesives to paperboard factories on a contract basis.

Constraints in starch-based adhesives production

The likely physical bottlenecks affecting this opportunity are access to cassava of sufficient quality for use in SBA production, and infrastructure problems. Another constraint is access to technical information required for development of locally produced SBAs that can compete with imported products by providing equivalent quality at a lower price. Currently Ghanaian manufacturers of SBAs compete well on a cost basis but their products are of inferior quality and that reduces their competitiveness.

Increasing prices for imported materials and a continuing fall in value of the Ghanaian Cedi against the US Dollar, have provided an impetus for Ghana's paperboard and plywood industries to look for cheaper locally produced alternatives to imported adhesives and extenders. Commercial adhesive manufacturers have shown that a reasonable product can be prepared from cassava and sold at a third of the price of the imported material.

The level of investment required to exploit this opportunity will depend on the approach adopted for production of the basic cassava chips. If a field-based approach is taken, each farmer will require a capital investment of approximately US$120 to cover the cost of a manual chipper, plastic sheet and 15 bamboo drying trays. Farmers would normally operate in groups of six, so the capital investment per group would be US$720. If a factory-based approach is adopted, US$700 will be required for each mechanical chipper. Artificial drying will increase costs by around US$10 000 depending on the design and origin of the drying unit. To prepare cassava flour for industrial use, approximately US$3 000 is required to purchase a suitable hammer mill and screening equipment with stainless steel screens covering the range from 1mm to 250um. To prepare premixed glue powders, an industrial mixer is required at a cost of US$8 000; alternatively using a cement mixer fitted with plastic shields to reduce dust could reduce costs. However, this is not a very satisfactory alternative and will increase hazards from chemical dusts in the factory environment.

The most likely economic bottlenecks affecting this opportunity are access to capital for processing equipment, and potential fluctuations in raw material prices and availability. Although cassava flour will be locally sourced, most of the chemical required for SBA production are likely to be imported and thus subject to supply problems, changes in market price and currency devaluation.

The production of cassava flour for industrial uses offers another attractive opportunity for cassava farmers who are willing to process cassava into high quality cassava chips. However, to realise this opportunity farmers will need financial and technical support to prepare chips and access to the market. This is likely to be provided by the manufacturers of the SBAs and glue extenders. These manufacturers need to be willing to provide the necessary financial and technical support as well as providing some form of infrastructure to maintain quality and provide a mechanism for transfer of chips from the field to the factory. The experiences of T&CG (Ghana) in the area of chips for livestock feed could prove invaluable in this case (see Chapter 7). It is evident from experience in Ghana that quality is an important issue for users of adhesives and glue extenders. It is therefore vitally important that the adhesive manufacturer meets the user's specifications at a competitive price. If these specifications are met, users will favour the local product because of the large cost savings made (typically 65 to 75%).

Locally produced SBAs based on cassava flour currently account for 45% of the market for paperboard adhesives. If manufacturers take more care over milling and screening of their products, and improve their understanding of the formulation of these products they could easily take over the entire market. The SBAs are essentially quite simple products and the necessary improvements are straightforward and could be achieved at low cost. However, many operators lack the technical knowledge to improve the quality of their products and have no access to external support.

In a conventional plywood process, synthetic adhesive contributes US$1.85-US$2.0 per sheet. Conventional glue extenders (maize starch & wheat flour) reduce this cost to US$1.38-US$1.54 (inclusive of cost of extender) depending on the cost and amount of extender used (Graffham & Dziedzoave 1998). If high quality cassava flour at a cost of US$65/tonne is used in place of wheat flour at a level of 25 kg of flour per 50 kg of adhesive, the cost of adhesive and extender would fall to US$1.32-US$1.40 per sheet. However, experience in the Philippines showed that an extender consisting of cassava starch mixed with sodium hydroxide, calcium hypochlorite and filler allowed the percentage of resin solids to be reduced from 45% down to 22% without loss of bond strength (Fidel et al. 1992). This formulation was able to reduce adhesive costs by more than 50%.

In 1998 trials with cassava flour, prepared at village level, from either peeled or unpeeled roots and screened to 250um using steel mesh have been successful. Two plywood factories in Kumasi carried out these trials independently and they yielded comparable results.

It was found that unpeeled root flour was not suitable for extending plywood glue; as it increased the percentage of reject boards from 1 to 70%. This was due to poor bonding resulting from high fibre and bark content (seen as non bonding blisters on the board surface).

In contrast it was found that peeled root flour could be used to replace 50-60% of wheat flour in the glue mix without loss of dry or wet bond (cold water soak) strength. When cassava flour was used at a substitution level of 50%, plywood production costs were reduced by 35%.

Industry representatives are very interested in this work. They are now interested in trials using flour based versions of the Philippines type product.

Plywood glue extenders offer an excellent opportunity for cassava flour, with a relatively large market of more than 2,300 tonnes per annum in Ghana. Cassava starch has been found to be superior to other starches for adhesive-manufacture because it can be used in its native form, is cheaper to manufacture, has good flow characteristics, forms stable pastes with neutral pH and is miscible with synthetic resins (Dux 1967). However, plywood manufacturers have been upset by the poor quality of locally produced products in the past. To exploit this opportunity, manufacturers of cassava flour need to provide high quality on a consistent basis to build up the confidence of potential users of the product.

A perspective on the Cassava starch industry in Brazil

Production

Cassava starch production in Latin America is located mainly in Southern Brazil. There are some small plants in Colombia, Bolivia, Argentina and Central America, but their production is minimal in comparison to Brazil. Lorenz, the largest Brazilian cassava starch production company, also owns part of a Venezuelan plant, and has made joint ventures with three plants in Paraguay, one with a daily root processing capacity of 400 tonnes and two with a capacity of 200 tonnes.

In 1997, Brazil's 80 production plants produced 300 000 tonnes of cassava starch. Most of the major plants are located in the States of Paraná, Sao Paulo, Mato Grosso do Sul (Table 3.8), while plants of lesser capacity are found in Santa Catarina State. Five or six small plants are established in other States, but their production is less than 3 or 4% of the national production. In total there are around 60 cassava starch companies in Brazil, many of which operate several plants.

Most companies produce native cassava starch. However, in the past five years, the number of firms producing modified starches has increased and production at the end of 1997 represented around 30% of total cassava starch production. Modified starches - cationic and acid starches - are used primarily by the paper and textile industries. Other modified starches include dextrin, maltodextrin and pre-gelatinized starches.

In late 1997, a cassava starch firm called INDEMIL started producing glucose syrup using an acid transformation process; thus diversifying its production. Currently, the company has a syrup production capacity of 80 tonnes a day. In 1998, INDEMIL through the "all enzymatic process" increased its hydrolysis production capacity.

Diversification within the cassava industry is inevitable, as more and more cassava starch firms penetrate into the traditional cornstarch markets, and as more cassava starch firms compete with multinational companies. The following section will examine the characteristics of cassava starch production, focusing on the main opportunities for competition in the starch market and the problems that the industry must resolve to realize these opportunities.

Characteristics of the industry

Brazil is a large consumer of starch, with an increasing demand for starch as production of starch using products continues to grow. This cassava starch market potential is matched by the potential to substantially increase cassava production, owing to untapped production potential that exists in many agricultural and industrial areas of the country.

One of the most important features of cassava starch within the starch market is its agricultural production potential. Three factors contribute to Brazil's potential to increase cassava production.

Good productivity: Between 1986 and 1998, the Brazilian cassava yields remained at around 13 tonnes per hectare. However, the situation in Southern Brazil does not reflect the national yields trends, because some farmers in the region produce up to 70 tonnes per hectare from 2 year-old cassava. Farmers in Southern Brazil, mostly of European origin, use modern production technology to achieve yields that are higher than the national average (Table 27) and are in fact among the highest yields in the world. In comparison, cassava yields in Thailand are only 14 tonnes per hectare from one year-old root.

High starch content: The average starch content in the cassava varieties cultivated in the State of Sao Paulo (Fiber, Branca de Santa Catarina, IAC 12, and Mico) were never below 30%, and were on average 33% annually (Sarmento 1997). In comparison, the starch content in cassava from Thailand is around 26%.

Availability of large areas: Small farmers on small land areas typically cultivate cassava. For a starch company, this means having to coordinate raw material supplies from a large number of suppliers. Cultivation on large land areas allows starch companies to reduce the number of suppliers and possibly make fewer contracts, enhancing the coordination of raw material supplies. In Brazil, this situation already exists. Many farmers currently cultivate more than 1 000 hectares of cassava and land is still available for further cultivation. Hence there exists a growth potential in the number of large farmers in the future.

Growth opportunities

The production process for making starch from cassava is very similar to that of making starch from potato or cereal. In fact, the production facilities needed to make cassava starch are even simpler because cassava is a purer starch source than potato or cereals^[6]. A cassava starch industry has excellent potential when compared with other cassava product industries for a number of reasons. Not only is there great agricultural potential but there is also the presence of large companies with access to excellent starch production technologies. The installation of a cassava starch plant in Brazil, with a capacity of 200 to 400 tonnes of root per day, costs between 1.5 million and 2.5 million dollars. The largest plants, with a capacity of 600 tonnes per day producing modified starches, can cost more than 10 million dollars.

The sales of an average Brazilian cassava starch company are between 2 million and 5 million dollars a year. Some of the companies reach 20 million dollars sales, and Lorenz - largest producing company in the sector - is said to have had sales of US$100 million in 1997, with 35% coming from exports (Venda de Sao Geraldo aquece a disputa pelo "Pao de queijo" 1998c). Lorenz specializes in modified starches, such as cationic, pre-gelatinized and acid modified starches, maltodextrin and dextrins.

Between 1994 and 1996, many cassava starch firms increased their production capacity. In 1997, the four largest Brazilian companies produced more than 20 000 tonnes of cassava starch each. Three of the four companies distributed production across several plants. The largest companies continue to modernize, gaining in efficiency and in economies of scale allowing them to compete more effectively with the modified cornstarch industries.

These companies, aware of this evolution, are in turn developing an interest in cassava starch production joint venture will allow a faster development of Lorenz, mainly in the Asiatic and North American markets. For example, National Starch entered into partnership with one of the eight largest Brazilian cassava starch companies, Halotek Fadel, and CPC International is looking for new partnerships. In May of 1998, another cornstarch company Mitsui established a joint venture with Lorenz, the largest Brazilian cassava starch company. This joint venture will allow a faster development of Lorenz, mainly in the Asiatic and North American markets.

A large number of companies are increasing the period of operation because of the need for a stable supply, rationalization of productive assets, better use of production capacity and a decrease in production costs. In the last three years, the annual period of operation significantly increased. Companies, located in the States of Paraná, Mato Grosso do Sul and Sao Paulo now operate 10 to 12 months a year.

As noted already, in Southern Brazil, cassava is gradually changing its status from a seasonal crop to a year round crop. The increase in production period influences the cassava root extraction rates. From May to August, a tonne of cassava typically produces 280 to 300 kg of starch.

This drops to 200 - 240 kg of starch from December to February. Brazilian cassava starch companies that operate year round obtain average rates of extraction of 25 to 27%, based on root weight. In the years 1997-1998, INDEMIL reached an annual average extraction rate of 30%, which represents so far the best cassava industrial productivity rate in the world. This company processed 140 000 tonnes of cassava in 1997 (two-thirds for starch and one third for "farinha"). These quantities prove that large production plants in Brazil are possible and are comparable to those already in Thailand. In 1998, INDEMIL had plans to process between 170 000 and 200 000 tonnes of cassava and to build another processing unit with a capacity of 1 000 tonnes of roots per day.

Industrial uses of starch/Major starch companies

The starch industry produces hundreds of products that are used in food and non-food markets of many different sizes. Starch demand can be classified in four main industries: food, paper, chemical and textile. Other industries that use starch though in smaller quantities include the oil industry, construction industry, etc. (Leygue 1994). In 1998, starch sales in Brazil totalled 1 million tonnes, of which 70% was produced from corn, and 30% produced from cassava^[7].

Currently, cassava starch is not imported into Brazil and exports do not exceed 20 000 to 30 000 tonnes. Cornstarch, on the other hand is imported into and exported from Brazil mainly by CPC International, but in small quantities. The Dutch company Avebe also imports potato starch to a small degree.

In general the starch market can be divided into three types: sugars, natives and modified starches, each one with a different potential for the use of cassava (Table 28).

In 1998, cassava starch competed favourably with cornstarch in almost every Brazilian market. In some markets, cassava starch exhibited clear advantages, and corn companies lost some of their position. The main markets for cassava starch are in delicatessens that use native starch, and paper and textile industries that use modified starches. Until the early nineties, modified cornstarches dominated the paper and textile industries.

However, modified cassava starches because of its superior properties now has a large market share in these two industries (paper and textile). The largest companies producing modified cassava starch for textile and paper markets are National Starch, a North American multinational, Lorenz, Halotek Fadel and INPAL. Other companies that are expanding include MCR and INDEMIL.

Constraints in the cassava starch industry

In spite of its potential a cassava starch industry faces a number of factors that limit its development.

Low level of mechanization in cassava cultivation: A comparison of production costs between cassava and corn reveals a different distribution of main cost items, (Table 29). The comparison indicates a higher use of additional technology such as inputs and automated activities in corn production as against cassava production. If mechanization were to increase in cassava cultivation, lower labour costs could significantly reduce cassava root price and allow economies of scale for cassava farmers.

Market concentration: Two companies in Brazil produce 700 000 tonnes of cornstarchs. In comparison, the cassava industry is made up of more than 60 firms. The sheer size of these two corn companies allows them greater opportunity to invest in product research, reach major customers and reduce production costs. All of which is more difficult for the smaller firms in the cassava starch industry.

Low education level: Large cornstarch companies typically employ specialists and professionals with good schooling for every kind of job. In contrast, in most cassava-based firms, the owners manage the company and make all the important decisions. In most cases, the owner may not have had the opportunity of higher education, which may negatively impact on the firm's growth opportunities. Another problem for smaller firms is the absence of technical sellers of starch, making it difficult to sell to large firms. As a result, cassava starch sales focus on a few traditional markets without expanding to new territory.

Lack of information: As a result of the relatively low education levels among managers and the small size of cassava-based starch firms, it is very difficult to access market and production technology information. The ultimate consequences, are difficulties in accessing new markets and technologies, price instability and problems caused by not recognizing new directions in future production.

Price fluctuations: Cassava prices are subject to great seasonal variations. They typically reach a minimum price level during the main harvest period (between May and August) and increase again by the end of the year (Figure 6).

Most agricultural raw materials are subject to seasonal variations in prices, both nationally and internationally owing more frequently to climatic and political factors. In Brazil, cassava starch competes directly with cornstarch. A comparison of raw material prices of both starches provides a good illustration of price fluctuations of cassava and cornstarch. In spite of great variations in corn price, the instability of cassava prices is a lot higher (Figure 7).

Strengthening the starch industry

Because bottlenecks do exist as reflected above, action is required to fully develop the cassava starch industry potential in Brazil. First, it is important to collect and disseminate market information such as price statistics for cassava and cassava starch within Brazil and around the world, as well as, the prices of main competing products. Equally, information on consumption statistics of starch within Brazil and around the world should be obtained and circulated. The main markets for cassava starch should be identified as well as the type of product in demand and in what quantities.

Technical information should also be disseminated about new techniques for producing different types of starches, e.g., native starch, modified starch and hydrolysed starch. Specialists should be involved. Managers should be educated about the different properties and uses of starches derived from different raw materials and through different processing techniques. Since many cassava-starch companies lack technical sales teams, technical sellers need to be provided and trained.

Secondly, another area requiring attention is the modernization of cassava production through the development of new cassava varieties, mechanized harvesting, and the development of improved cultivation techniques and machines. The raw material cost of cassava roots represents between 50 to 60% of all costs within the cassava starch industry. This includes production costs and explains the importance of reducing cassava root costs. Other costs in the starch industry are wages and value-added taxes, each representing 13% of all costs. The remaining costs include electricity at 4%, packing at 3%, and repairing and wood at 2% each, (Vilpoux 1997).

Modernizing of cassava production should help to stabilize prices and supplies. Today, many processing companies are regulating supply and prices by vertically integrating with cassava producers. It is argued that improved cooperation between companies and farmers, as well as the fixing of minimum prices and planting and harvesting periods, would be more efficient than vertical integration (Vilpoux 1997).

It is also argued that the price of final starch-based products could be reduced by improving the starch extraction rate. For example, it is currently possible to achieve an extraction rate of 90% starch from root, with an end product of 12% humidity and 3.5% other components. The productivity in commercial starch will be 35% of the total weight of the root. At this rate, processing could improve significantly from its current rate of 25 - 26% in Brazil and Thailand.

These needed improvements can be facilitated by governmental institutions, such as research centres and universities, IAC (Sao Paulo State), EPAGRI (Santa Catarina State) and IAPAR (Paraná State) for agronomic researches, and CERAT/UNESP (Sao Paulo State) for technology and market analysis. Nevertheless, even without scientific help, the cassava starch industry can and will continue to modernize and resolve its problems. The advantage of government intervention is that it could accelerate the modernization process and allow smaller companies to access the evolution.

Points of interest

Analysing future trends within the industry makes it possible to focus on a few important points:

Increase of industry concentration: Industry concentration has been increasing in recent years and should continue to increase in the future. The concentration ratio or percentage of national production of the four largest cassava starch producers was 27% in 1995 and rose to 35% in 1997.

The concentration ratio of the eight largest cassava starch companies has grown from 44 to 55% over the same time period (Vilpoux 1998). Although production at the largest cassava starch companies in Brazil continues to increase, it is still lower than the average in Thailand, which is 30 000 tonnes of starch per year. Multinationals, like CPC International and National Starch, continue to show interest in cassava starch production in Brazil. National Starch, for example, is involved in a joint venture with a Brazilian company. Lorenz also recently made a commercialisation agreement with the Japanese company Mitsui. Other European multinationals, as Cerestar, Roquette frères and Avebe continue to study the possibilities of starting production units in Brazil. As the Brazilian cassava starch industry continues to concentrate, many existing companies could be closed within the next few years.

Modernization of cultivation: Cassava producers are using more and more herbicides, allowing a decrease in cassava production costs. The harvest, which is mainly done manually and which represents the main production cost, is becoming more and more mechanized, allowing economies of scale and cost reductions to take place. In future years, more harvesting technologies will continue to become available.

Modernization of industrial process: Extraction rates are increasing very fast, allowing for increased productivity. In 1993, the average annual extraction rate was 25%, and in 1998 the average increased to about 26-27%, with one company reaching 30%. Meanwhile, the period of production continues to increase towards all year production and at full capacity.

Modernization of cultivation and processing should reduce costs and make cassava starch more competitive with cornstarch. Table 30 illustrates, the competitive situation of the Brazilian cassava starch industry in relation to producers in Europe. The Table shows that only potato has a better starch productivity per hectare, though a shorter processing time, which handicap production and increase total costs.

A cassava root price of US$35.00/t (possible to reach with the adoption of new technologies in big areas and an extraction rate of 30%), allows the sale of starch at a price of about US$250.00/t. This is competitive with cornstarch prices in the United States and lower than starch prices in other Latin American countries. Such a price would allow an expansion of cassava starch production in Brazil and exportation of the product to other countries, not only in South America but also in the United States and Europe.

This work has focused on the potential of cassava as a cheap source of starch. Its advantages are relatively high yields of between 20 and 25 tonnes per hectare and good extraction rates of over 30% starch to total root weight. Mechanization will allow further reductions of root crop prices and will stimulate increased acreage and supply to starch firms. Cassava planting has already been mechanized but the most complicated part is harvesting. In 1998 it was possible to find farmers cultivating more than 1 000 hectares of cassava. However, yields are relatively low because of the large quantity of labour required. Yields should increase in the near future.

A factor limiting the expansion of the cassava industry is the great variation in raw material supply that influences the price stability of cassava and its end products. In Brazil, in spite of this instability, cassava starch prices remained competitive in comparison to corn, with an average price of US$350.00/t from January 1990 to December 1997. Currently, the largest companies are attempting to stabilize production and guarantee farmers a minimum price of US$36.00/t to stimulate cassava production. These companies are also regulating the supply with their own production of cassava. For example, in 1996, in the main producing regions, cassava starch companies obtained on average between 20 to 30% of their total cassava needs from their own production (Vilpoux 1997). The biggest firm however, which utilized 140 000 tonnes of roots in 1997 (for starch and "farinha"), does not produce cassava anymore, preferring instead to buy its raw material supplies with and without contracts.

Modernization, growth, and the development of new technologies will strengthen the industry in the near future. Multinational companies such as National Starch and CPC International are already very interested in the evolution of cassava starch production. International investments in cassava could allow Brazil to reach Thailand's similar levels of development and supply Latin American markets. The competitive advantage of Brazil is its high cassava productivity and the root starch content.

The installation of large cassava starch plants in other Latin American countries, such as Colombia, Bolivia, Venezuela and Paraguay are possible and some already do exist in Venezuela and Paraguay. Industrial and agricultural requirements are similar to those discussed for Brazil. The bottleneck for the installation of these production plants is the distance from main consumption markets. Starch is a product that is consumed by other industries and its consumption increases with overall economic development. In Latin American countries, cassava consumption regions are located in Southern Brazil, Argentina and Chile. Since cassava is easily grown in only the first region, Brazil has therefore the opportunity to be the main cassava starch producer in South America.

A perspective on the cassava starch industry in Thailand

In Thailand, the cassava starch industry is export-oriented. The industry is rapidly shifting from the export of native starch to the export of modified starches. According to the traders, modified starch is broadly classified into two major categories: physically modified starch and chemically modified starch. Monosodium glutamate (MSG), derived from native starch and a major user of native starch in Thailand, is not treated as a modified starch.

The Thai modified starch industry

The Thai modified starch industry began in 1984. By 1998, there were 49 cassava starch factories, 11 making modified starch or equipped with modified starch facilities. Most are joint ventures with developed countries. Of the 49 factories, 26 factories are located in the Northeastern region, while Nakhon Rachatsima and Korat regions have altogether 11 factories. The remaining 12 factories, are located in the Central region (Eastern seaboard) and in the Northern region. The Thai estimated annual production capacity is 1.8 million tonnes of native starch, of which 300 000 to 500 000 tonnes are further processed into modified starch.

The native and modified cassava starch industries are export-oriented. From 1993 to 1997, total export volumes of native starch increased from 460 000 to 872 000 tonnes. The average export price increased from 5 070 baht/tonne to 7 140 baht/tonne (US$200 to US$275). Taiwan is Thailand's largest importer of starch, among a group of 10 major importing countries. The quantity exported to Taiwan increased from 180 000 tonnes (39 percent of the total exports) in 1993 to 313 000 tonnes (36 percent) in 1997. Japan is the second largest importer. The quantity exported to Japan, however, has decreased from 114 000 tonnes in 1993 to 76 000 tonnes in 1997. From 1994 to 1997, Indonesia imported more than 150 000 tonnes per year. This contributed significantly to Thailand's increased native starch exports (Table 31).

Exports of modified starch also increased during the 1993-1997 period. In 1993 and 1997 modified starch exports were 195 000 tonnes and 264 000 tonnes respectively. Among the 10 major importing countries, Japan imported more than 55 percent of Thailand's modified starch exports. The quantity exported to the Netherlands and Indonesia also increased, especially to Indonesia. The quantities to the other seven major importing countries have been fluctuating. The markets for Thailand's native starch and modified starch are rather concentrated and limited (Table 32). Seven of the ten major importing countries of modified starch are also the major importers of Thai native starch.

In the 1988-1997 period, the export price of native starch (f.o.b. Bangkok) fluctuated between 4,398 baht/tonne in 1989 and 7,922 baht/tonne in 1995 (US$173 to US$308). A similar pattern was experienced in the price of hard pellets. During the same period, the export price of modified starch increased every year from 4,683 baht/tonne to 8,736 baht/tonne to 15,475 baht/tonne (US$185, US$348, US$611, respectively). The one exception was in 1992-1993 when a slight decrease was observed. During the period 1988-1997, the modified starch price was on average about two times higher than native starch, and four times that of hard pellets (Figure 8).

Factors underlying Thai modified starch industry growth opportunities

At least two favourable physical factors contributed to these growth trends. One factor is the increased accessibility to improved infrastructure. This makes it possible to set up factories in more remote areas where land prices are low and pollution problems can be addressed with low cost technology. A second factor is the availability of domestic mechanical engineering firms, equipped with improved mechanical know-how and the ability to duplicate new imported technologies, to expand factory capacity at a comparatively lower cost. As a result, native starch production increased while the number of factories remained almost constant. In fact, Thai firms have set-up native starch factories for firms in Indonesia and China.

An important economic factor that has contributed to the growth of the export market has been the comparatively lower trade barriers imposed on the imports of modified starch by the importing countries. Another economic factor that contributed to the growth of the export market has been the high costs of domestic modified starch in some developed countries, such as Japan. Therefore, the joint ventures with developed countries provided not only technology transfers to Thailand, but opened the export markets for Thai modified starch.

Constraints

Two factors limiting the expansion of native starch factories are land prices and pollution. Town expansion, urbanization and the boom in real-estate (housing and recreation) in most of the cassava growing regions are the main causes of inflated land prices and complaints about the pollution produced by the native starch factories. The development of modified starch is slowed by the lack of commercially viable modern domestic technologies for modified starch production. Thai research on modified starch technologies and utilisation exists, but the results of this research have not yet been tested and adopted. Currently, most of the modified starches are produced using imported technologies provided through joint ventures.

The bottleneck for export growth is the existing subsidy on corn and potato starch in the developed countries that are the major markets for modified starch. These subsidy policies also jeopardise the export growth of Thai modified starch into new markets, because Thai modified starch has to compete with the subsidised low-price modified starch exported from the developed countries.

In the domestic market, the initiation of domestic modified starch production opened up the possibility of import substitution in Thailand. This was particularly true for those modified starches for industry usage e.g. paper and packaging industry. In addition, the expansion of the Thai agro-industry has opened a potential market for modified starch in both food and non-food sectors. In general, one can envisage that there is a sizeable domestic potential demand for modified starch, which has yet to be penetrated through an active technical sales promotion.

The limitation to domestic market expansion is that prospective industrial users lack knowledge of the availability of modified starch. In addition, prospective industrial users lack confidence in using the domestically produced modified starch as a substitute for imported modified starch in their production line. This is owing to the fact that the input cost of modified starch constitutes only a small proportion of the final cost of most industrial goods. Therefore, the saving realised from switching from imported modified starch to the locally made starch is not sufficient to compensate the risk of having problems on the production process.

The Thai flour industry's associations

Native starch factories are the stakeholders in the development of modified starch. The recorded numbers of factories increased from 50 in 1970 to 146 factories in 1978. But in 1989, only 47 factories were registered as members of the Thai Tapioca Flour Industries Trade Association (TTFITA). The estimated annual production capacity was 1.3 million tonnes of native starch. By 1995, the number of factories registered with the TTFITA increased to 54 factories with an estimated production capacity of 1.4 million tonnes per year. Three years later, the number of registered factories decreased to 49, but with an estimated annual capacity of 1.8 million tonnes. During this time span, the location of native starch factories had shifted from the Central region (Eastern seaboard) to the northeastern region, now the major cassava growing area (Table 33).

Native starch factories have improved and expanded their production capacity and implemented both forward and backward integration. After setting up modified starch operations, some factories exported their own products while others sought joint ventures with exporters and foreign firms. Others preferred renting out their premises to other firms for the production of modified starch. All factories played an active role in the development of native and modified starch and were involved in the development of government price intervention programmes. These factories also participated in the dissemination of new improved varieties to cassava growers and initiated contract-farming systems between cassava growers and starch factories.

More than 80 percent of the total registered members of the Thai Tapioca Flour Industries Trade Association are also exporters of both native and modified starch. Thai native and modified starch industry exporters have access to market information on each importing country and can explore new market possibility for the starch industry. In most cases, exporters are the initial contact point for bringing foreign joint ventures into the modified starch industry.

The trader associations have explored new market opportunities through trade missions. These missions have been assisted by concerned government agencies. The trade associations and exporters have also been involved, to some extent, in the formulation of cassava policies and in setting up cassava research and development agendas, through their participation in ad hoc committees set up by the government.

The Thai Tapioca Trade Association (TTTA) and the Thai Tapioca Flour Industries Trade Association (TTFITA) are the two most active private institutions involved in policy, research and development of native and modified starch. The close linkage between these two associations is due to the integration between the respective industries. These associations have, to a certain degree, determined the research and development agenda of the Thai Tapioca Development Institute Foundation (TTDIF).

The TTDIF was founded in 1992 with an endowed fund of 600 million baht, collected from the auction of EU export quota in 1991. The Foundation has six major activities: 1) promotion of research and development on selected new varieties of cassava, 2) carry out research and development on (cassava) production structure, 3) addressing problems of tariff and trade barriers, 4) evaluation of the efficiency of competitive exporters, 5) establishing information units within TTDIF that can assist the public and private sectors, and 6) carrying out activities with non-political charitable organisations that are beneficial to society. (The Thai Tapioca Development Institute 1998).

Examples of the Thai Tapioca Development Institute Foundation's funded research include the reduction of cassava production cost by using high yielding varieties such as KU 50 and Rayong 90, market research and news activities, and technology development such as improving the efficiency of starch factories. The Foundation's funded research also focused on new high value-added products (bio-degradable products) and animal feed, information collecting and disseminating, and human resource development, such as overseas short training, seminar and exchange researcher programs.

There is a sizeable domestic market potential for modified starch. In 1995, native starch used for producing modified starch was estimated at 490 000 tonnes. This increased to 512 000 tonnes in 1997 of which 61 percent was used in producing modified starches (both physical and chemical); 25 percent was used in glucose and fructose syrup production; and 8 percent each was used in producing sobital and high fructose. Total starch utilisation is expected to increase to 583 000 tonnes by 2002. It should be noted that the estimation was carried out before the current financial crisis. The estimated native starch utilisation showed that sobital and other modified starch had relatively high growth rates that reflect increasing potential domestic demand for modified starch (Table 34).

Since 1993, Thai exports of modified starch has increased every year at an annual growth rate of 8 percent. During 1993-97, Japan was by far, the largest market for Thai modified starch with an estimated annual growth rate of four percent. There is an increasing export market potential for Thai modified starch, especially for industrial use in Asian countries. Modified starch is mostly for industrial use, and cassava modified starch has to compete with modified starch from corn and potato.

Based on the current growth rate, modified starch exports from Thailand in 2002 could be 388 000 tonnes. With the devaluation of Thai baht against the US dollar, this export quantity could be more.

Lessons learnt

The lesson learned from the development of modified starch industry in Thailand can be summarized as follows:

Size of export market: The large scale of the Thai export market encouraged farmers to grow cassava rather than other cash crops. Farmers who are able to produce cassava with high starch content could sell roots for native starch processing, while those producing roots of low starch content could sell roots for chips and pellets for animal feed.

A native starch industry: Development of a stable native starch industry provided the basic material for further processing into high value-added products and modified starches.

Free markets: A relatively free market environment contributed to the development of both the native and modified starch industries.

Mature base of industries: The co-operation of government agencies allowed the mature industries for native and modified starches, as well as, those for animal feed industry, to organize, set up and institutionalize research and development in support of sustainable development within the industry.

The Thai success in export market validates a number of old adages: firstly one must recognize and seize opportunities when they arise, and secondly success begets success. A number of countries in the sixties and seventies would appear to have been well- positioned to export cassava products, but they never realized the opportunity.

Future directions

The case studies of domestic cassava starch markets provide a consistent picture that there is a potential, often times unexploited market for cassava starch. Modified cassava starch is becoming increasingly important, but the predominate cassava starch is still native cassava starch. In general, domestic cassava starch is competitively priced to imported starches. The examples of Benin and Malawi, representative of many smaller cassava-producing countries, illustrate the problem of' having a growing market for starch, which is unfortunately still very small. The current small size of the starch market in many countries discourages the development of "modern" starch factories. For example, the Malawi market can only use about twenty percent of what could be considered to be a "traditional" cassava starch plant and Benin only three percent. Even in Ghana, the market requirements are approximately what can be produced by one traditional cassava starch plant working four to six months in a year.

The case studies of Thailand and Brazil illustrate that cassava starch can be competitive in the world market and in domestic markets. Thailand exports about a million tonnes of starch, about one third of which is modified. Thailand also utilizes about 800 000 tonnes of starch (native and modified) domestically. Brazil's domestic market is approximately 300 000 tonnes. About one third of Brazil's starch is produced from cassava.

Whether the domestic cassava starch industry is small or export-based oriented, there are a number of barriers to its development. The most commonly identified barriers, noted in the case studies, are a lack of consistent supply of raw material, unfamiliarity with the potential users of cassava starch and poor quality control. Furthermore small-scale starch processing tends to limit production of modified starches, increasing concern over the cost of cassava.

A number of the case study countries felt that they could export cassava starch. However it must be recognized that their cassava starch production capacity, or proposed capacity, is less than one day of Thai cassava starch production. In addition, all countries, including Thailand, may encounter price support regimes in importing countries (especially developed countries) that are barriers to exports.