Part: 4.The economics of briquetting
Contents - Previous - Next
18.Introduction to the economics of briquetting
Chapter 19.Capital costs
Chapter 20.Operating costs of briquetting
Chapter 21.Total costs of briquetting
Chapter 18.Introduction to the economics of briquetting
The financial costs of briquetting are very dependent upon the nature of the project, in particular upon the raw materials used and the plant location. The object of this section is to provide an indication of what range of costs can be expected for both operating and capital costs and what is the resulting unit product cost. Whether or not briquetting is economic in any given location will depend critically upon how these unit costs relate to the price of the likely substitute fuels. Such prices vary so widely that it is impossible to generalise about the likely economic benefits of briquetting. It is hoped that the costs set out here will enable some preliminary decisions to be made as to the likely role of briquetting in particular circumstances.
The first two sections deal with capital and operating costs whilst the final section looks at the total unit costs of briquetting.
Chapter 19.Capital costs
The capital cost of a plant is not always easy to establish on a consistent basis.
In different circumstances there may be different conventions about what is included in capital cost: for example, estimates made by funding agencies tend to include a fairly generous amount for initial spare parts whilst commercially funded plants often include spares as part of operational costs.
Another major source of variation is whether the plant is a stand-alone operation or whether it is part of an existing plant. In the latter case, it is often possible to utilise existing buildings and to eliminate a significant element of capital cost. It is not unusual in plants with dedicated new buildings for these together with site civil works to equal the cost of new machinery.
Probably the biggest variation in plant capital costs, however, comes from the raw material to be used and the form in which it is collected. There are three reasons for this.
First, the nature of the residue may be such as to significantly lower the rated output of the briquetting machine. Normally machines are rated according to their output using wood-wastes and if less dense residues are used then the actual throughput may be less by as much as 50%. In one Indian plant visited, the machine used was capable of delivering 1.5 tonnes/hour fed with sawdust but only 1.0 tonnes/hour using rice-husk. As the cost of a briquetter seems to be almost linearly dependent upon rated output, the unit capital costs associated with a wood residue are likely to be lower than for most agro-residues.
Secondly, the form of the initial feed may require more or less rigorous treatment before it is fed to the briquetter as such. Some wastes, for example sawdust, often need to be dried to reduce their moisture to the 15% or so tolerated in most machines. The cost of a drier can easily equal or even exceed the cost of the briquetter itself. For example, the Brazilian manufacturer, Biomax, which makes both presses and driers, quotes a list price ex-factory of US$39 000 for a 1.1 tonnes/hour piston press, US$11 000 for associated conveyance and silos, and US$52 000 for a dryer plus heater of l tonnes/hour capacity. It is likely that in this case the cost of the briquetter is unusually low but the dryer is clearly a big additional cost.
Other ancillary equipment may include various types of chipper or shredder to reduce the size of the residue. This will be required for various forms of wood-waste other than sawdust and for certain types of agro-residue such as cotton stalks.
Thirdly, there is a very big difference between field-residues such as cotton-stalks or straw and factory-residues such as sawdust or coffee-husk if the cost of the equipment to collect the residue is included in the initial capital cost. This is not always the case; the straw-briquetting operations which have been set up particularly in Germany, often assume that existing tractors and trailers are available to collect the straw from the field.
If this is not the case then the capital costs of collection equipment are likely to be large. In a recent study of the costs of a wheat-straw briquetting plant in Ethiopia (World Bank 1986), it was estimated that 3 straw-balers, 5 trailers and 6 tractors would be needed to move 5 000 tonnes/annum of straw. The estimated cost was US$107 500 before transport to Ethiopia.
Finally, it should be noted that one important aspect of capital cost is what may be called the engineering and design standards of the plant.
This is difficult to define but easy enough to appreciate when making a site visit. Many, though by no means all, of the commercial plants visited were cramped, very dusty and sometimes dangerous with unguarded fly-wheels and rudimentary wiring. Such plants are often the ones which are squeezed into residue-producing plants to utilise existing buildings. Naturally, the capital costs of such plants are lower than plants which have firm floors, proper electrical fittings and are reasonably spacious.
This aspect is not necessarily related to the use made of labour rather than machinery to convey and store the residues and product. However, it is likely that in a plant where little or no use is made of mechanical handling, there will be a need to have spacious buildings to avoid problems of dust and dangerous overcrowding. In one of the Indian plants visited, this meant that the capital cost of the new buildings used was almost equal to the cost of the machinery. Thus savings made on ancillary handling equipment in favour of human labour may be partially illusory if reason able working conditions are to be maintained.
Nevertheless, it is probable that some of the development-agency funded projects are over-designed relative to privately funded projects with too great an emphasis placed upon mechanical handling, storage and bagging.
In order to put some numbers on these general comments, Table 12 shows the capital costs of six plants in various countries. Three of these are design studies (Plants A,B,E) which contain rather detailed cost breakdowns whilst three are of actual plants (C,D,F) for which rather less detailed breakdowns were available.
Table 12 shows the estimated capital cost per tonne of output for each plant assuming that it is required to pay back the initial capital sum with interest in a 10 year period. Three interest rates are shown: 7%, 10% and 15%. A number of other conventions could be used for the capital charge but they would make little difference to the results.
It can be seen that the plant with the lowest capital cost is Plant C based upon a unit at a wood-working factory in Brazil. This may be regarded as an extreme case; the raw material is dense and needs only chipping to provide an acceptable feed whilst the cost of a piston-briquetter in Brazil is much lower than anywhere else. The cost of buildings was minimal, amounting to little more than some covered space alongside the factory building. The products were removed daily to local customers so little on-site storage was required. The only equipment required in addition to the briquettes was a chipper and some simple pneumatic conveyance to a feed silo. The capital cost of 4.3 US$/tonne at 10% interest can be regarded as very much a minimum for any operation.
The four plants D,E,F,A are interesting as despite being for different materials at different sizes and in different countries, they show very much the same level of capital costs. It should be borne in mind that the figures in Table 12 have been put into a common currency and therefore may suffer from the usual problems of multi-currency comparisons.
Plant D is based upon an actual operation in Kenya where a small screw briquetter has been installed in an existing plant to feed an on-site wood boiler. The only equipment needed additional to the press was some modifications to an existing dryer. The amounts shown for spares and engineering and installation are notional as the expenses were not specifically itemised.
Plant E is a design study (World Bank 1987) for a new sawdust-based plant in Ghana using 6 Taiwanese screw-presses which would be supplied with a dryer. The additional equipment includes a lorry to collect raw materials and a fork-lift vehicle onsite.
Plant F is an actual Indian plant using rice-husk feed. The material requires no preparation and, as all handling is manual, there are no additional equipment costs. This low level of equipment costs is probably one reason why rice-husk briquetting has managed to retain some commercial viability in India despite its apparent disadvantages of high inherent ash content and raw material costs.
The plant required completely new buildings and these were provided, at this plant, on a generous scale. In other Indian plants, old buildings are used with a substantial cost-saving.
Plant A is a design study (World Bank 1986) for a plant using coffee-parchment sited at an existing plant in Addis Ababa which is planned to move in a short time. It is rather highly engineered with pneumatic conveyance throughout and a large intermediate storage silo. At the new site, the building costs are estimated at over US$57 000. Coffee parchment requires no preliminary treatment before feeding to the briquetter so the additional equipment costs shown are all for handling and storage.
The final plant, B. is a design study (World Bank 1986) for an operation based upon straw collected from the fields in Ethiopia. The very large additional costs required for this kind of unit are obvious with the capital charge at 10% rising to nearly US$32/tonne. The extra equipment needs arise not only from the field collection equipment but from the bale breaker, hammer-mill and associated conveyance equipment.
These investment costs, although high, are not out of line with the capital costs reported for straw-briquetting in Europe. One plant (KTBL 1983) of 1 000 tonnes/annum throughput is quoted as costing US$150 000 in 1983, the equivalent of about US$900 000 when adjusted to the 500 tonnes/annum throughput of the proposed Ethiopian plant in 1986 prices. Another plant of only 375 tonnes/annum throughput is quoted as costing US$90 000 for equipment plus US$35 000 for storage equipment; the equivalent of about US$1.7 million when adjusted to 5 000 tonnes throughput. Although these German costs are difficult to compare directly on a consistent basis, they suggest that the briquetting of field residues is a very expensive operation which can only be justified in a context of high fuel prices. Straw briquetting is also justified in Europe because of environmental constraints on the burning of straw in open fields.
It seems reasonable to suggest on the basis of these figures that for centralised residues in most situations, a capital cost of 9-12 US$/tonne at an interest rate of 10% is reasonable. In circumstances where the provision of new buildings can be cut to the very minimum this might be reduced a little. Higher interest rates, and in most developing countries commercial loans would be somewhat higher, would add to this figure. A 15% rate would raise the level to about 11-14 US$/tonne.
Table 12: Comparisons of Capital Costs
|(all in '000 US$)|
|Transport and delivery||16.3||54.1||4.5||0.0||12.3||0.0|
|Annual output||5 000||5 000||4 000||300||3 500||4 000|
|Capital cost assuming a 10-year finance period|
|(US$ per tonne)|
Plant A based upon World Bank coatings for a plant in Ethiopia
at an existing coffee plant using 1 piston and 1 screw machine
Plant B based upon World Bank coatings for a plant in Ethiopia at at new site using 2 piston machines
Plant C based upon list-prices for a Brazilian plant at a wood-factory using 1 Brazilian piston machine
Plant D based upon a Kenyan unit at an existing factory using 1 screw machine and a modified existing dryer
Plant E based on World Bank coatings for a plant in Ghana using 6 Taiwanese screw machines with dryers included.
Plant F based upon an actual Indian plant using 1 Indian piston-press with no ancillary equipment located in new buildings.
Chapter 20.Operating costs of briquetting
The labour costs of a briquetting plant are very dependent upon plant design; the scope of the operation, in particular whether there is any significant residue collection activity; the wage rates of the various categories of labour employed; and the extent to which the unit is integrated with a larger factory which can supply some labour needs on a part-time basis.
There is a tremendous variation possible between plants depending upon the balance of all these factors. It is, for example, quite possible to run a briquetter unattended for long periods, even overnight. This is common practice in Sweden where high wages place a considerable premium on labour reduction and where a wood-waste plant can only be profitable if left to operate automatically. Naturally, such operation increases the capital cost greatly and is not considered here as an appropriate mode in any developing country. However, without going to such extremes, it is possible to observe wide differences in the labour use in different operating plants.
The least direct labour use can be seen in plants which are in-house operations, that is plants contained in larger factories either to briquette their own waste or to produce briquettes for their own boilers.
In these situations, the labour costs of administration, loading and maintenance may be wholly or partly absorbed in the general activities of the factory. Thus in one Kenyan plant, only one unskilled worker was paid to work directly on briquetting; other labour costs were contained in costs for plant overheads or maintenance. Even so, the attributed labour cost was still quite high amounting to rather more than 5 US$/tonne of product.
The size of plant seems to play a significant part in determining labour costs with larger operations having unit labour costs up to 50% less than those of small plants. In India, a small plant based upon the carbonising technology appeared to have costs of about 8 US$/tonne without taking into account any salary to the owner/manager. At a single piston-press plant, labour costs were 3 US$/tonne including payments to 4 managerial staff (2 involved in selling the product), 15 unskilled workers employed seasonally and 4 skilled workers employed full-time.
An operational plant in the Sudan, working one-shift per day, employs 10 people including 2 operators and 1 trained engineer. The remainder are unskilled loaders, packers and guards. The smaller number of people employed as compared with the Indian plant may be related to the greater labour need of moving rice-husk as compared with groundnut shells. However, there is normally a wide variation found in the numbers of unskilled workers employed at such plants.
The total wages bill in the Sudanese plant was calculated as being Sú13.4/tonne of product which was 12% of the total production cost. Although the Sudanese exchange rate has charged too fast in recent years for comparisons to be reliable, this is equivalent to 1.2-5.5 US$/tonne of product depending upon whether a current or the historic rate is used. No managerial wages were included in this Sudanese estimate nor anything for maintenance engineers. One might expect a somewhat higher wage cost if these were to be included.
As a general rule, one would expect the labour costs for a large plant (that is one producing in excess of 3-4 000 tonnes annually) to be in the range 3-5 US$/tonne provided the residues were centralised at the plant. If they had to be collected from the fields then labour costs would rise sharply. There are no data available on actual operational plants but a study made for a proposed Ethiopian plant (World Bank 1986), based on maize residues, proposed to employ over 250 people on collection and storage of the residues. The very low labour costs in Ethiopia kept the labour component down but, even so, total labour costs for the plant rose to 16.7 US$/tonne.
There is a strong trade-off to be made between labour and investment in mechanisation for any field-residue collection but it must be expected that, under any regime of mechanisation, labour costs would exceed 10 US$/tonne. In German plants based on straw, with high labour costs and correspondingly high levels of mechanisation, the labour cost can exceed 30 US$/tonne of product.
Despite the wide variations in labour cost, briquetting as such is not labour intensive relative to the unit capital costs and the cost of maintenance, power and other consumables. Labour costs seldom exceed 15% of the total and are usually much less. Even in the maize-residue plant, mentioned above, the estimated labour cost was only 15.4% of the total estimated factory costs including an annualised capital charge.
The maintenance of briquetting machines can be quite onerous particularly for screw presses and for piston presses operating on abrasive materials such as rice-husk. The actual assignation of maintenance costs varies according to the internal accounting practices of plants and to how maintenance is undertaken.
In some operations, in particular those using screw presses, an operative may be employed virtually fulltime on building up worn screws. Such work, though effectively continuous maintenance, may be registered as a labour cost. Even in such cases, however, the cost of welding rods may be high.
An analysis of the various plants visited as well as design studies and manufacturers' data suggests that the maintenance costs of piston presses are likely to be in the range 3-8 US$/tonne, whilst for screw machines, the equivalent range is 512 US$/tonne. This latter figure is wide partly because of the differences in treating maintenance costs but also because there are clearly variations between machines.
It is apparent that maintenance costs can be a significant element of briquette production costs and one that is often underestimated in planning plants. It is a factor which may place a continuing reliance on imported spare-parts possibly leading to delays in production if adequate allowance is not made in advance.
Most of the plants discussed here are powered by electricity from a mains supply. There is no reason why plants in remote areas should not use diesel generators (as does the Sudanese plant quoted above) or use direct drive from diesel or steam engines.
Analysis of available data suggests that for a practical plant, including some allowance for power-driven conveyance equipment, power consumption will fall in the range 50-80 kWh/tonne of product for piston presses and 100-120 kWh/tonne for screw presses. The unit power consumption of screw presses is quoted for small machines of about 150 kg/hour capacity and lower figures are quoted by the manufacturers of larger machines. These may be an accurate representation of the power savings available as machines get larger. However, there are relatively few of such large screw presses installed and good operational data are sparse.
Power prices vary considerably between countries; a range of 5-8 cents/kWh probably covers most places though in, for example, Brazil, power prices are much lower.
At such prices, the power costs of briquetting could vary between 2.5 US$/tonne and 6.4 US$/tonne for piston machines whilst small screw machines would have higher unit costs of up to about 9 US$/tonne.
The diesel-powered Sudanese plant consumed fuel which cost 19.2 Sú/tonne of product or 7.6 US$/tonne at the exchange rate then prevailing. As diesel prices move closely with the exchange rate this is probably a reasonable dollar estimate of the real fuel cost. This suggests that diesel powered plants will have costs at or somewhat above those for electrically powered units.
It is common for the operators of plants located inside factories to have no clear idea of their power costs as electricity bills may be spread across other power-consuming units. The costs of power use do not seem to be regarded as a commercial problem warranting any attention and most plants put their unit power costs in the range of 2-5 US$/tonne. Some plants, particularly in Brazil, give even lower estimates but Brazilian electricity is very cheap.
The price paid for residues, if any, is very much a site-specific issue and must be regarded as an add-on unit cost relevant to a particular project. However, even if the residue is nominally free, it is common for a transport cost to be incurred in bringing the residue to the briquetting plant. In the case of units sited at the residue production point this is avoided but for other situations the transport cost may be a significant part of operational costs.
As with maintenance, transport costs may be added into general labour costs if fulltime drivers are employed. A more common situation is to hire appropriate transport as a single briquetting plant needs to be rather large to warrant a dedicated lorry and driver.
In India, the cost of bringing rice-husk from local mills by bullock-cart was reported to be a little more than 1 US$/tonne whilst the cost of fetching wood-residues and sawdust in Brazil varied between 3 and 6 US$/tonne depending upon the distance moved. The latter figure corresponds to quite large transport distances and would be an unusually high level.
These include taxes, if appropriate, insurance, selling costs, consumables such as lubricating oil, packaging if needed and so on. It would be unusual if these dropped below 1 US$/tonne for any plant.
Chapter 21.Total costs of briquetting
Almost all the cost categories discussed above depend to a more or less significant degree either on accounting conventions (this is particularly relevant for the calculation of capital charges) or on site or country-specific factors. In addition no allowance has been made for the cost of raw materials over and above transport charges. The country studies make it clear that where briquetting has become commercially viable to a degree there is a tendency for residues to acquire a market price where previously they were free.
The cost ranges derived above for a large piston machine are:
|Other||at least 1|
A simple addition of the least and greatest costs would suggest that a piston-machine briquetting plant would have total factory costs in the range 20-36 US$/tonne of product. It would however be misleading to adopt costs in the lower part of this range except under the most favourable circumstances; these might be the use of dry wood-waste drawn from the immediate locality in a country where labour costs and power prices are low and where fairly low-cost machinery is available.
A possible location meeting these criteria is Brazil; there a company planning to set up a number of large briquetting plants in the interior has suggested that total costs would be about 26 US$/tonne including some payment for wood-wastes. This must represent very much the bottom end of the cost range.
In other, less favourably situated countries, it is much more likely that total costs would be towards the upper part of the range. It should be emphasised that these do not include any allowance for residues being priced nor for any profit element.
It would be expected that the unit costs for screw presses would be somewhat higher than for piston machines. They do not appear to offer any significant advantages in investment costs and in some cost categories, notably maintenance and power, they are likely to be more expensive. They also appear to have higher unit labour costs though this is probably a factor relating to scale of production rather than any intrinsic feature of screw presses.
The higher intrinsic costs of the screw machines may however be offset by the fact that their small production levels, and indeed small physical size, means that they can be, literally, squeezed into low-cost situations. These would typically be a small wood-plant able to site a machine right by the waste pile in a building which needs little or no modification.
One Kenyan user who has put a small screw press into such a favourable situation (except that it utilises residue from a nearby sawmill) has assessed total factory costs at about 21 US$/tonne including depreciation and finance. This includes no allowance for raw material transport costs and may underestimate power and maintenance costs. If these are corrected then it is likely that true factory costs are more like 25 US$/tonne.
In general therefore, it would be wise to assume that total costs for briquette production are in excess of 30 US$/tonne and may be above 35 US$/tonne except in particularly favoured circumstances.
These numbers are in accordance with the situations in both Brazil and India, the two developing countries where briquetting has managed to establish some kind of commercial basis. In Brazil, it seems possible to survive by marketing briquettes somewhere above 30 US$/tonne whilst in India a marketed price in excess of 40 US$/tonne is required. In both cases, these prices produce bare commercial survival rather than large profits. In India, it is common to pay up to 15 US$/tonne for rice-husk; in Brazil, wood-wastes are usually cheaper if charged at all.
These broad cost levels refer only to plants based upon factory residues. The costs for any field residue plant will be much higher.
It is clear that in many countries, the price of fuelwood is well below these levels to an extent that effectively rules out briquettes as commercial propositions. This is often true even if it is assumed that industrial consumers are prepared to pay a premium for briquettes as they are of consistent and reliable quality.
There are exceptions to this. It is reported (World Bank 1986) that fuelwood prices in Addis Ababa reached 83 US$/tonne in 1985; even allowing for a retail markup this allows considerable scope for briquettes to undercut fuelwood. However, industrial fuelwood prices in, for example, Kenya do not exceed 20 US$/tonne, a level which briquettes cannot hope to reach. Similar low fuelwood prices in Thailand have effectively destroyed the local briquetting industry despite very low investment costs and good raw material availability.
It is probably true that in most countries, the current level of fuel prices is too low to justify briquetting. Nevertheless the examples of Brazil and India show that it is quite possible for fuel prices to rise to levels where briquetting is viable. Such a rise can, moreover, take place quite rapidly: the fuelwood price in Addis Ababa was reported to be only 9 US$/tonne in 1973.
The key policy issue in many countries with respect to briquetting must be the judgement as to the extent to which it is worth supporting briquetting activities, in spite of their immediate lack of profits, in expectation that fuel prices will rise in the future. There is no general answer possible to this. There are many countries where fuelwood prices seem set to remain relatively low for the foreseeable future. Indeed in some countries there is genuine optimism about the possibility of stabilising prices indefinitely by the development of fuelwood plantations.
However there are also countries where it seems likely that deforestation must cause a rise in prices in the not too distant future. In such circumstances briquetting of agro-residues can have a legitimate economic role without any need for subsidies.
Contents - Previous - Next