Cover Image
close this book Local Experience With Micro-Hydro Technology
close this folder F. ECONOMIC CONSIDERATIONS
View the document 1. BASIC APPROACH a) Cost-Benefit-Approach for Socio-Economic Selection
View the document 2. MICRO-HYDROPOWER AND LARGER HYDROPOWER PLANTS
View the document 3. MICRO-HYDRO PLANTS AND OTHER ALTERNATIVES

2. MICRO-HYDROPOWER AND LARGER HYDROPOWER PLANTS

a) Experience with Tangible Internal Costs

b) Experience with Tangible Internal Benefits

c) Experience with Tangible External Costs and Benefits

d) Experience with Intangible Costs and Benefits

 

All figures given from examples, and experience with hydropower plants are to be taken as order of magnitude since every case will differ very much for reasons of different labour costs, equipment-cost, site-cost, import-cost etc. Furthermore, figures reflect different cost/revenue-years; it would be academic to de- and revaluate foreign currencies to a comparable international statistical basis in U.S. Dollar ($) since neither inflation-rates nor rates of exchange can accurately be secured.

a) Experience with Tangible Internal Costs

The World Bank states that the economic limit for mini-hydro projects (under 1 MW) is in the $ 2'000 -3'000 per kW range. One should consider this statement as too rough and a cost per kW of $ 3'000 as too high to electrify hamlets, micro-industries and villages, in view of experience gained in a number of projects. A report from OLADE 1'000 to 2'000 per kW as the desirable goal, and still other sources state that efforts should be made to remain close to $ 1'000 per kW for obvious reasons. Based on this, it is now of interest to look at some actual figures, to see how far this can be achieved.

A series of different types of turbines from European manufacturers is shown in fig. 75, in the output range from 30 kW to 300 kW, and a head range from 2,3 m to 350 m. The costs given are updated to the level of 1980 and include complete generating equipment (e.g. turbine, step-up transmission where necessary. flywheel, governor, alternator, valves and other acessories, but excluding penstock). It becomes clear from the table that equipment for low head and low output becomes very costly.


Fig. 75: Different Hydro Generating-Equipment Costs

Source: Integration GmbH, Laufwasserenergie

The same source states that euqipment cost ranges from 40 to 50 % of total cost in conventional hydro installations of the sizes referred to. This means that -since low-head installations have a relatively larger flow rate - civil construction costs amount to 50 % of total cost for heads above 20 m, and to 60 % for lower heads. This is of course broadly generalised but it serves the purpose of arriving at a relevant magnitude of total cost.

Based on fig. 75, the calculated total costs are between $ 1'825 and $ 8'750 per kW for heads from 2,3 to 13,5 m and $ 1'000 to $ 3'000 for heads between 27 and 350 meters. Thus a very clear cost-function of head and size is reflected in total plant costs. Notwithstanding this trend, it is also true that equipment for the highest head with an output far above the average of the examples given, is not the cheapest. Much depends on the type of equipment and the suppliers pricing. It is nothing unusual to get quotations for the same site, which differ by several hundred percent from different suppliers. This adds another variable, making representative cost analysis still more difficult.

As far as costs of civil construction-components are concerned, no standard cost unit can be given here. Dams, canals and intakes will obviously cost a very different share of the total for different sites. Much depends on the topography and the geology, and also on the construction method applied and the materials used.

Of interest in the context of this paper is now a comparison with costs of local technology. The examples in fig. 76 are all actual project costs at about the price level of 1980. In all cases shown, the turbine is of local design and construction. Other equipment components are also made locally, such as step-up transmission, flywheels, coupling, base frames, gate-valves and penstock.

Governors are not used at all in the smaller plants, while on the bigger sets, different solutions were found as indicated. Alternators are in all cases imported; either from within the region or from overseas.

Comparing data from fig. 75 and 76 brings to light a number of facts. These should not be interpreted to be the absolute and only truth for reasons of enormous project diversity. However, the trend clearly stands as shown:

· The influence of variations of head and size on price are not pronounced because of a high degree of flexibility as to the chosen equipment configuration, which is appropriate to the situation.

· The average share of equipment cost of the total is 26,5 %, e.g. clearly less than the conventional 40 to 50 %.

· Total costs for the range from 10 to 100 kW using local technology, compared to cost per kW in fig. 75 show reversed economies of scale.

· Taking the average figure for equipment cost of $ 265.-/kW and comparing this with the average of $ 890.-/kW of eight turbines (imported in fig. 75, shows that on an average, locally made sets (fig. 76) cost approx. 30 % of imported equipment only. (To arrive at a more representative average cost for imported equipment, the first three figures - representing atypical cost because of low head/low output -have not been included).

· Taking amounts for total cost for both series of examples shows, that only with local technology can the stated goal of $ 1'000 per kW be approached under "normal" conditions.

 


Fig. 76: Cost of Generating Equipment Using Local Cross-Flow Turbine

Sources: BYS, Nepal; NEA, Thailand; ITB, Indonesia

 

Conclusions:

· It is possible to counter the rule of traditional economies of scale by using local technology in the range up to 100 kW.

· It is also possible to reduce traditional overall costs considerably.

· The result is, that the range up to 100 kW -where local technology is possible -becomes more economical as compared to the range from over 100 kW to 1 MW.

This latter point can not directly be derived from the examples shown but is the result of an evaluation of hydropower activities in Nepal, in the range up to 500 kW. The study referred to maintains that there are two economically feasible ranges of size in evidence. The first from 2 to 30 kW ( and the second from 1 to 3 MW because in the latter, economies of scale are pronounced. For the range from over 100 kW to 1 MW, it is the following factors that make it uneconomical:

-too big for local technology, therefore high capital cost

-requires skilled and professional staff, therefore high running costs

-too small for a remarkable effect of economies of scale.

The same authors also introduce a new dimension of cost-reduction for marginal village electricity-supply: Many small hydropower installations exist in Nepal for mechanical power supply to agro-processing units, such as mills and oil presses. These projects are operating profitably, and could therefore be equipped with small alternators to provide domestic lighting and very limited other electricity uses, at only marginal cost of $ 250 to 420 per kW (which may be extended to 100 kW if head is over 20 meters as shown in fig. 75) this also includes electricity distribution). As indicated previously, one can take it for granted that the upper limit of these very economical hydropower plants can be extended from 30 kW to about 100 kW, without jeopardising the intangible benefits of still using appropriate technologies, which are locally manageable.

Similar cost-experience has been gained with micro-hydro installations in China. The cost-concept is based on the conviction that canals, dams, roads etc. can be attributed to irrigation and flood-control anyway, so that electric power can be calculated at extra-cost for the generating equipment and civil works implied. In this way, costs attributable to electricity generation and distribution of as low as $ 350 per kW, are reported.

As to the distribution of energy, local low-tension lines (380 V) were included in the first-level cost outlay. L.T.-lines are possible up to a couple of kilometers, depending on the material used for wiring. Good examples exist in Nepal or in Tanzania. Most of the Tanzanian micro-hydro plants have a short overhead distribution line to the consuming points, with an average length of 800 meters (min. 200 meters, max. 2 km). The electrical energy produced by these micro-hydro plants, is used mostly for households, water pumping, productive and agricultural machinery, hospital-stations, etc. With such short distances, a low-tension line is sufficient. Since the local low-tension distribution network must be provided anyway, the cost of local distribution would remain the same whether the electricity-supply comes from the local powerstation, or from grid extension fed by a distant power plant. But the point is that the villages have practically no chance to get a grid-extension, since step-down transformation on the high-tension side is a costly thing. One must rather envisage a grid-extension in a maybe not so distant future, after the local power plant has contributed to the village development, so that this can afford later-on to integrate its consumers into a larger grid-system.

Another cost-component of distribution is the metering of energy-consumption. It depends much on the tariff applied and will be discussed later.

To conclude on tangible internal cost, the operation-and maintenance-cost shall briefly be sketched: Hydropower plants are characterised by high initial capital-investment and low operation and maintenance cost, whereas diesel-powered generators are cheaper in terms of investment, with high running fuel-cost. For hydropower plants, operation cost must be seen as a function of the size of the plant and the salaries of local staff. Maintenance cost depend more on the characteristics of the site (rebuilding of intakes, removing slides on canals etc.) than on the size of the plant. Experts on this problem maintain that operation and maintenance cost for small hydropower plants (between 100 -400 kW) are almost independent from installed capacity and amount to roughly $ 25'000 - 40'000 per annum. Consequently, operation and maintenance cost per unit produced, decrease rapidly with size. They conclude that operation and maintenance costs can only be paid by the revenues of the plant, if its capacity is higher than 200 kW, the load factor is at least 20 % (with an availability of 75 %) and at high electricity prices. This again fosters evidence that the small-sized plant between 100 -1'000 kW should be avoided.

Smaller plants than 100 kW show a drastic reduction of operation and maintenance cost by a factor averaging 4 -5. However, the tendency to overstaff also micro-hydro stations is a problem. It also seems that the running-cost of a micro-hydro plant cannot easily be covered by simply selling electricity at a too high rate; the solution will rather be an adequate shaping of the tariff, in order to promote use of electricity to obtain a good load-factor.

A further element of the operation and maintenance cost is the depreciation of the plant. Again experts have calculated a weighted rate of depreciation, according to the different lifespans of the plant's parts like civil-work, generating-equipment etc. They suggest an annual depreciation rate of 4 % of the initial investment. Usually this determines such a high price per kWh in the range of up to US {1 for depreciation alone that a serious problem is posed. Form an economic point of view one has to take a firm stand regarding depreciation. As long as the perpetuum mobile is not invented there is no such thing that does not depreciate. If depreciation of the plant cannot be paid, then the country's development budget or foreign agencies must replace the whole installation, after the wearing out of the plant. If depreciation can be paid, two ways are conceivable: A loan can be repaid step by step thus diminishing capital-cost of the plant, and proportionally increasing the new debt-potential for the new plant, or, internal reserves are accumulated, so that the total replacement of the plant can be paid with proper financial means.

In most cases a subsidy to the plant will be inevitable, be it for private or public hydropower investment. No one would ever question the fact either that schools, streets and other public services have to be depreciated at the public ' expense. However, to what extent costs are covered, is entirely determined by the revenues, a question that will be dealt with in the following chapter.

b) Experience with Tangible Internal Benefits

Information thereof will focus on the prices -and tariff - situation whereas "surplus revenues" and the "IER" are not elaborated at length.

"Baliguian" in a very remote place in the Philippines, is an example of locally designed and locally built equipment (Welded Francis turbine). With a head of 27 meters, and an installed capacity of 100 kW, the hydropower plant caused total cost of $ 749 per kW (excluding distribution cost), of which hardly 40 % for the generating equipment ($ 287). Producing 437'500 kWh per year - implying a load factor of close to 50 % -the selling price is an amazingly low US-cents 2.1 per kWh (calculated from data in: Dumol, Mini-Hydro Application, p 31 Looking at investment cost, it appears that capital interest and depreciation are not accounted for). Other small and micro-hydro installations in the Philippines sell at higher prices:

Installation

US-cents per kWh

-Magat A & B with 2'800 kW capacity

4.7

-Agua Grande with 2'750 kW capacity

5.2

-Hasaan with 30 kW capacity

5.3

The third example may demonstrate once more that a micro capacity is by no means in itself a guarantee for a low total investment and selling price. There are so many variables which finally determine the economy of a plant, that relying too much on the cost element of the generating equipment, can be a misleading yardstick.

The East-African power plants mentioned earlier, sold at rates per kWh between US ¢ 13 and 18 in 1966, a not so attractive offer considering the fact that European electricity producers calculate today with a price of US 3.5 per kWh to cover cost and depreciate the plant within 10 years (see Integration GmbH, Laufwasserenergie, p 55). At US 4 2.0 per kWh the depreciation is stretched over 20 years.

Back to China -a country with a lot of experience on small hydropower plants -more comforting evidence is available. It is said that micro-hydro plants sell cheaper electricity than larger plants; in case of higher investment-cost per installed kW in small plants, this is by far offset by the high transmission costs of larger plants. For the Xinhui county, a price of US 5 per kWh is indicated for small hydel stations and US ¢ 8.4 per kWh for grid-electricity (1979). It is certainly true that kWh-prices in developing countries often turn out to be higher than in industrialised countries; but the cost of alternative energy sources often are even higher, or -if lower as in the case of wood - have intangible costs which are unmeasurable. At this point, one should simply remember the question of energy-source selection from a thermodynamic standpoint. From this angle, cooking with electricity for example, would energetically be a luxurious undertaking.

Uneconomical and thus unwanted uses of generated power can be controlled by shaping the tariff accordingly. Any electricity supply company has to deliver electrical energy at the moment of production. Due to the fact that electrical energy is not storeable, an electricity supplier has to charge for two services: for keeping a certain amount of generation equipment ready (which is a typical public utility duty), and for the actual supply of energy per period of account. There results a two-part tariff, one part for the power-capacity, the other part for energy consumption. For hydropower plants with relatively high capital investment and low running cost (no fuel-cost), this means high capacity and low energy charges, whereas thermal generation induces moderate capacity and high energy charges. The two-part tariff however requires the metering of the individual energy-consumption, as well as the consumers peak demand. If upto 90 % of total consumption is by small consumers, this is simply too expensive.

An alternative method is based on the assumption that the power requirements of a domestic consumer correspond to the number of rooms or bulbs etc. he possesses. The tariff applied is then a maximum-demand equivalent charge per room or per electrical device. Metering is not anymore necessary. The method is called the flat rate per unit installed. This tariff does in fact -though in a very simple way - also take into account the two part cost structure of an electric energy supplier: when the electric device is switched on, the flatrate charged to the consumer pays for energy consumption; when it is turned off, the flat-rate pays for capacity installed. The rate is in both cases the same. The negative aspect of the flat rate is of course that it induces consumers to have the electrical devices on as much as possible, instead of saving energy. However in the case of hydropower, this aspect should be looked at from another angle: the price per kWh can only decrease, when the generating equipment is utilised to the largest possible extent. The electric energy supplier is therefore interested in having a high load factor. The flat-rate tariff is in fact a consumption-inducing tariff.

Another, also consumption-inducing tariff and thus load-improving method is to price the supply in blocks at a decreasing rate, e.g.

-the first 1'000 units at 10 US ¢ per kWh

-the next 2'000 at 8 U ¢ per kwH

-all additional units at 6 US ¢ per kWh.

The method increases the load factor and lowers the average price per unit of energy. This can also be combined with an off-peak tariff, thus flattening out uneven capacity use. However, both measures favour the larger consumers.

Finally, a fourth method should be mentioned: the tariff of the social untility-principle. To every use a different use-value is attached, expressing the users' appreciation of the different services rendered by electricity . For example, lighting receives a high use-value, whereas the energy-input into workshop machinery will be rated lower for reasons of its productivity. To further exemplify: The "Mel river" power plant in China charges the following rates:

type of consumer

US ¢ per kWh

-Industry

1.2

-Domestic

2.1

-Irrigation and pumping

0.6

 

The "Gu Don Mountain" power plant in China puts the social utility weight differently. It looks at industry as the most solvent partner, since precisely industry -by means of electricity - is enabled to attain surplus revenues, and is therefore charged most, the domestic-sector second-most, but hardly anything for irrigation and pumping (though surplus revenue will be generated in agriculture in the first place).

Consequently, the tariff helps determining many things like the load factor and thus the average price per kWh, the peaks and off-peaks in consumption and the socially wanted distribution of the generating costs, by differentiating the tariff according to end-uses. But all in all, the tariff will have to provide an acceptable IER to the producing organisation.

c) Experience with Tangible External Costs and Benefits

As to external costs, some data will be given here specifically focusing on H.T.-grids as the most crucial factor, whereas state subsidies and foreign-exchange costs are not further quantified.

If one looks at the Brasilian hydroenergy potential of 209'000 MW -14 % of it which is utilised only - the energy problem of this country seems to be solved. This "global" optimism is an illusion however, because the distances from the sites of hydropotential to the main consumer points are enourmously long. And so are the implied costs of H.T.-lines.

Considering a transmission-line of 10 kV, the Chinese indicate cost per km of $ 5'500.-to 6'500.-. Very similar costs are budgeted for a H.T.-line of 7 km length, including transformation to 6 kV in Nepal, with $ 6'250.- per km (1977).65) The World Bank stated in 1975 that extending a subtransmission link 25 km to an isolated demand point, may cost around $ 100'000, thus amounting to $ 4'000 per km.66) Inflation will bring it close to the above-mentioned $ 6'000.- per km. In Tanzania* a typical price for a 33 kV line is approx. $ 8'430.-per km.67) A rough average -at costs of today -is therefore approx. 7'000 $/km. This is valid only in the small-hydro range.

Some analyses have been done of a comparison between public supplies from the main grid and micro-hydro stations.68) For micro-hydro -serving local consumers and public lighting - an installed capacity of 80 kW was considered, and public supplies from the grid was assumed to be of medium-voltage of 33 kV as subtransmission link to the local L.T.-line. The capital costs of supplies from the grid are much higher than those of micro-hydro generation, but are of course depending on the distance from the nearest grid. The crucial point however, is this: the higher the utilisation of the energy-project, the better off is the supply from the public grid, since its fixed transmission costs result in a decreasing unit-cost, if the load-factor increases. This is illustrated in fig. 78, where unit-cost is given for different situations. At a low level of utilisation, public supplies from the grid are too expensive because it is extravagant to extend so expensive networks at $ 7'000.-per km, to meet small demands in areas remote from the grid. Microhydro thus is better off, provided that its local consumers are not too remote either from the power plant, since -as shown earlier - L.T. lines are technically restricted to a relatively small radius of some kilometers at the most. As soon as it comes to transformation and H.T.-lines, micro-hydro tends also to run into high costs.

When the socio-economic uplift of the village has increased energy-consumption, the marginal internal cost of expanding micro-hydro generation, will have to be compared with the external cost of integrating the village into the main grid. Again the Chinese have even applied a combination, by using small hydro up to its capacity-maximum (maximising the load factor) and supplementing peak-demand by public supplies from regional grids.


Fig. 78:Average Costs of Different Schemes, US $ per kWh

Source: World Bank, Rural Electrification, p 21, micro-hydro figures are calculated on the basis of project data from Nepal

Some brief remarks should be added to the tangible external benefits. Whether tax revenues will rise depends on surplus revenues generated by the higher energy input into production, be it agricultural or industrial. Product diversification and possible price decreases cannot be generalised here. Similarly, the question of subsidies must be looked at from case to case; it is conceivable that subsidies for imported kerosene will turn into subsidies for depreciation of a micro-hydro plant. But it might as well be that subsidies increase, but can be activated in terms of accounting, through local build-up of innovation-centres, more local employment, less expenditures on the trade-balance of the country, etc. It is especially the import bill, which could substantially be affected by implementing micro-hydro plants.

Most of the civil work will be entrusted to local engineering firms thus accounting for 30 -50 % of the total cost. One has of course to safeguard against the tendency of importing steel, cement etc. used in too elaborate construction. As to the mechanical and electrical components, what is possible by local means has been elaborated in chapter D at length.

d) Experience with Intangible Costs and Benefits

As indicated earlier this third-level assessment is the most difficult one. The problem will be demonstrated by means of few examples without aiming at a comprehensive description.

Measuring the very long-term discharge of a river over 10 -40 years can be a cost which nobody really considered. The tropical and subtropical zones have instable climatic features which can play costly tricks on a hydropower plant. The intangible cost of not having correctly assessed the hydro-potential of a chosen site, will suddendly turn into very tangible costs once the water discharge is substantially lower in a dry year. In China the general reliability of micro-hydro plants is rated lower than the one of large power stations. Severe droughts periodically plaguing large areas of China can incapacitate small stations quite rapidly.69) Therefore an intangible cost can consist in the interruptions of energy supply from small power stations, causing some disorder and planning problems to the economic life of a village. But experts working with micro-plants insist that -if capacity is based on minimum-flow rather than a power plant capacity aiming at optimum-utilisation of flow - such incidents are seldom and when they occur, the harm done is not so dramatic.

Certainly one does also have to consider some socio-cultural impacts of a new energy-source like electricity. Examples are: smoke from open fires in houses projects inmates from flies and insects; electrical appliances do not. A fire also heats a house, an electrical cooker does so much less; the fire offers light in the mornings and evenings and provides a natural center for social life, whereas electrical appliances do not substitute for these functions.

On the intangible benefit-side in the first place one has to evaluate the immense "mobilisation-effect" of rural electrification, all the development thrust which can be generated in a small community that makes suddenly power available for irrigation and drainage, for primary processing tasks such as grain treshing and milling, fodder crushing, oil extraction, timber sawing etc., truly the first steps towards sensible modernisation in remote, poor communities. What counts are not only the tangible external benefits (surplus revenues) of the above-mentioned activities, but above all -intangibly -the motivations, the attitude and the drive towards "being able to do something" against poverty.