ARTICLE IN PRESS

Energy Policy ] (]]]]) ]]] ]]] www.elsevier.com/locate/enpol

Comparison of options for distributed generation in India

Rangan Banerjee,1
Department of Engineering and Public Policy, Carnegie Mellon University, Pittsburgh, PA 15217, USA

Abstract There is renewed interest in distributed generation (DG). This paper reviews the different technological options available for DG, their current status and evaluates them based on the cost of generation and future potential in India. The non-renewable options considered are internal combustion engines fuelled by diesel, natural gas and microturbines and fuel cells red by natural gas. The renewable technologies considered are wind, solar photovoltaic, biomass gasication and bagasse cogeneration. The cost of generation is dependent on the load factor and the discount rate. Gas engines and Bagasse based cogeneration are found to be the most cost effective DG options while wind and biomass gasier red engines are viable under certain conditions. PEM Fuel cells and micro turbines based on natural gas need a few demonstrations projects and cost reductions before becoming viable. A strategy involving pilot projects, tracking of costs and dissemination of information is likely to result in DG meeting 10% of Indias power needs by 2012. r 2004 Elsevier Ltd. All rights reserved.
Keywords: Distributed generation; Annualized life cycle cost; Load factor

1. Introduction The earliest electric power systems were distributed generation (DG) systems intended to cater to the requirements of local areas. Subsequent technology developments driven by economies of scale resulted in the development of large centralized grids connecting up entire regions and countries. The design and operating philosophies of power systems have emerged with a focus on centralized generation. During the last decade, there has been renewed interest in DG. This paper reviews the different technological options available for DG, their current status and evaluates them based on the cost of generation and future potential. The relevance of these options for a developing country context is examined using data for India.

Different denitions of DG have been proposed. Some have linked this to the size of the plant, suggesting that DG should be from a few kW to sizes less than 10 or 50 MW. Ackerman et al. (2001) provides a review of alternative denitions of DG and suggests that DG be dened as the installation and operation of electric power generation units connected directly to the distribution network or connected to the network on the customer site of the meter. DG is also referred to as dispersed generation or embedded generation. DG options can be classied based on the prime movers usedengines, turbines, fuel cells or based on the fuel source as renewable or non-renewable. There are a large number of possible system congurations. In this review the comparison is limited to the following options: (A) Non-Renewable 1. 2. 3. 4. Internal combustion engine fuelled by diesel Internal combustion engine fuelled by natural gas Micro-turbine fuelled by natural gas Proton exchange membrane (PEM) fuel cell with reformer fuelled by natural gas

Corresponding author. IIT Bombay, Energy Systems Engineering,

Powai, Mumbai 400076, India. Tel.: +91-22-2576-7883; fax: +91-222572-6875. E-mail address: rangan@me.iitb.ac.in (R. Banerjee). 1 On leave from Indian Institute of Technology Bombay. 0301-4215/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.enpol.2004.06.006

ARTICLE IN PRESS
2 R. Banerjee / Energy Policy ] (]]]]) ]]] ]]]

(B) Renewable 5. Wind turbine 6. Solar photovoltaic (PV) 7. Biomass gasier connected to a spark ignition engine (dedicated gas engine) 8. Bagasse cogeneration in sugar factories Other options that have not been considered here are small hydropower, geothermal, ocean thermal, tidal and solar thermal power generation options. In order to place DG in the context of the Indian power sector, a brief background of the Indian power scenario is provided before comparing the DG options.

2. Indian power sector India had an installed capacity of 105,000 MW (Ministry of Power, 2003a, b) in the centralized power utilities on 31st March 2003. Of this 74,400 MW is accounted for by thermal power plants, 26,300 MW of large hydro plants and 2700 MW of nuclear. The focus of power planning has been to extend the centralized grid throughout the country. However the capacity addition has not been able to keep pace with the increasing demand for electricity. This is reected by the persistent energy and peak shortages in the country. The transmission and distribution losses are extremely high (estimated to be more than 25%, this includes theft). India has a plan to add 100 000 MW of additional power generation capacity by 2012 (MOP, 2001). This requires an average capacity addition of more than 10,000 MW per year. Centralized generation alone is unlikely to meet this target. In this context DG is likely to be important. DG also has the advantage of improving tail-end voltages, reducing distribution losses and improving system reliability. The present installed capacity of DG is about 13,000 MW (10,000 MW diesel, 3000 MW renewables). The majority of this is accounted for by diesel engines that are used for back-up power (in the event of grid failure) and operate at very low load factors. The share of the energy generation from DG is marginal (about 23% of the total generation). Apart from the diesel engines, the DG options that have been promoted in India are modern renewables. India is probably the only country with a separate Ministry of Non-conventional Energy Sources (MNES). The renewable energy installed capacity was 205.5 MW in 1993 (104.6 MW small hydro, 39.9 MW Wind). This increased to 2978 MW in 2001 (as on 31st March 2001) and accounted for almost 3% of Indias installed power capacity (MNES, 2001; Annual Reports MNES, 2000, 2001, 2002). The growth rate of installed renewable power capacity during the period 19932001 was 39% per year. During the period January 2000April 2001

the installed capacity increased from 1600 MW to 2978 MW (an annual growth rate of 49%). Fig. 1 shows the installed capacity of different renewable energy technologies (Annual Report MNES, 2002). The major contributors are small hydro o25 MW which accounts for 1341 MW (45%) and wind which accounts for 1267 MW (42%). The installed capacity in Biomass based power generation is 308 MW (10.3%), with most of it coming from bagasse based cogeneration. Most of the installed capacity available from renewables is accounted for by grid connected systems (wind, small hydro and biomass cogeneration). This accounts for about 3% of Indias installed capacity contribute to about 12% of the total generation (due to low capacity factors on renewables). The growth rate has been signicant (above 30% per year). This has been facilitated by an enabling policy environment and a supportive government. Despite the emphasis on extending the centralized grid to the rural areas, 78 million rural households (Ministry of Power, 2003b) or 56.5% of rural households are still unelectried. The recently passed Electricity Act (2003) has made it a statutory obligation to supply electricity to all areas including villages and hamlets. The act suggests a two pronged approach encompassing grid extension and through standalone systems. The act provides for enabling mechanisms for service providers in rural areas and exempts them from licensing obligations. MNES has been given the responsibility of electrication of 18,000 remote villages through renewables. The ministry has set up an ambitious target of meeting 10% of the power requirements of India from renewables by 2012 . In most cases, the areas to be electried do not have sufcient paying capacity. Most systems are subsidized by the Government or the utility. The power sector has signicant losses and needs to ensure that the DG systems selected are likely to be cost-effective. This paper examines the cost effectiveness of the different DG options selected.

1600 Installed Capacity (MW) 1400 1200 1000 800 600 400 200 0 Wind Small Hydro Bio Combn Bio Cogen Gasifiers Waste Energy Solar PV 63 210 35 15 47 1267 1341 Total Renewable Installed capacity 2978 MW 31/3/2001 MNES

Fig. 1. Installed capacity of renewables in India.

ARTICLE IN PRESS
R. Banerjee / Energy Policy ] (]]]]) ]]] ]]] Table 1 Input cost data used for calculations Option Diesel engine Gas engine Micro turbine Fuel cell Capital cost (Rs./kW) 25000 33000 45000 141000 Life 20 20 20 10 Z Efciency 40% 35% 28% 45% O&M cost Rs./kW h 0.25 0.25 0.25 0.25 3

Discount rate d= 0.1; natural gas price = Rs. 5200/1000 sm3; diesel price = Rs.16/l, density = 850 kg/m3; LHV = 9700 kcal/kg. Sources: ICRA, 2003; Borbely and Kreider, 2001; Resources Dynamics Corporation, 2001.

3. Comparison methodology
Annualised Life Cycle Cost (Rs/kW/year)

50000 45000 40000 35000 30000 25000 20000 15000 10000 5000 0 0 0.2 0.4 0.6 Load Factor 0.8 1
Diesel Gas Engine MicroTurbine Fuel Cell

In order to compare the costs of generation of electricity from each of these options, the annualized life cycle cost (ALCC) is used. The annualized life cycle cost represents the annual cost of purchase and operation of the system. The cost of generated electricity is obtained by dividing the ALCC by the annual generation. The ALCC is computed as ALCC C 0 CRFd ; n ACf ACO&M ; 1

where C 0 is the initial capital cost for the option, ACf is the annual fuel cost for the option and ACO&M is the annual operating and maintenance cost for the option. The capital recovery factor (CRF) is computed based on the discount rate d and the life of the option n using the equation CRFd ; n d 1 d n =1 d n 1: 2

Fig. 2. Comparison of annualized life cycle costs for non-renewables (Discount rate = 10%).

The annual generation is dependent on the load factor. The cost of generation is dependent on the size and the application load factor. In this paper a 100 kW peak rating is used as the basis except for wind and biomass cogeneration that are considered to be in the range of a few MW. The calculations are done with existing Indian fuel and equipment prices. In the case of technologies not commercially available in India the existing international prices in US $ have been converted to Indian rupees at the prevalent exchange rates (1 US $ 47 Rs. in 2003). An idea of the comparative costs of options and impact of the load factor will provide an idea of the viability of the DG option. The status of each option in India is discussed along with some of the issues relevant for its adoption.

4. Non-renewable cost of generation Table 1 shows the input data used for the economic calculations. Fig. 2 shows the annualized life cycle costs of the diesel, gas engine and micro-turbine options, as a function of the load factor. It is clear that except at

very low load factors, the gas engine and microturbine option seem cheaper than the diesel engine. One of the main reasons for this is the availability of relatively cheaper natural gas (Rs. 0.144/MJ of energy) in India as compared to diesel (Rs. 0.464/MJ of energy). The ratio of the diesel price to the natural gas price on a per unit of delivered energy is 3.2. In the US the price of natural gas in January 2003 (USDOE, 2003) was 4.47$/1000ft3 (Rs. 0.167/MJ) and the price of diesel oil to industrial consumers was 82.5 c/gal (Rs. 0.297/MJ) resulting in a ratio of 1.8 of diesel price to natural gas price. It is likely that the differential between diesel and natural gas prices in India would reduce in the future. This comparison is done with a societal discount rate of 10%. The price of power from diesel engine generators is Rs. 4.8/kW h (10 c/kW h) at 80% load factor with fuel cost accounting for 86% of the cost of generation. Fig. 3 shows the cost of generation from diesel engine-generators as a function of the load factor. The industrial tariff prevalent in Maharashtra is shown for comparison. It is seen that electricity from the grid is cheaper for load factors greater than 15%. (Tariff

ARTICLE IN PRESS
4 R. Banerjee / Energy Policy ] (]]]]) ]]] ]]]

45000 40000 Annualised Life Cycle Costs Rs/kW installed 35000 30000 25000 20000 15000 10000 5000
Fig. 3. Cost of generation from diesel engine-generator.
Diesel Gas Engine Micro Turbine

0 0 0.2

for large industrial consumers receiving high tension supply.) For private sector companies that have a higher discount rate1 d 30%, the comparison is shown in Fig. 4. It is seen that diesel engines are preferred at load factors of 20% or less (less than 5 h per day). This is probably the reason for the large base of diesel enginegenerators for back up power in India that provide uninterrupted power supply in the event of grid failure. Diesel engines are manufactured indigenously (major companies include Cummins, Wartsila, Kirloskar and Greaves) and there is signicant experience in India in the operation, maintenance and repair of diesel enginegenerators. About 10,000 MW of diesel engine capacity exists in India. Only a small portion of this is connected to the grid. Most of these operate with very low load factors. From the national viewpoint there is an attempt to discourage diesel based power plants since India has a middle distillate bulge (scarcity of middle distillates like diesel) that is constrained by the renery mix and necessitates the import of petroleum products. Natural Gas engines are not as common, probably because natural gas was not available around the country and the higher initial capital cost. The improvement in natural gas availability and the presence of gas distribution companies is likely to see an increase in gas engines. Microturbines are not indigenously available. A joint venture between Allied Signal and Thermax was announced, but was subsequently discontinued. For fuel cells, the technology considered here is the PEM fuel cells (USDOE, 1998) that operates at low temperatures 80  C. The disadvantage is that it can only withstand a small proportion of impurities (carbon monoxide). Fuel cells are not indigenously available
1 Private companies in the manufacturing sector India perceive a scarcity of capital and have high discount rates. The bank interest rate in India has recently been reduced and it is expected that this would result in a lowering of company discount rates.

0.4 0.6 Load Factor

0.8

Fig. 4. Comparison of non-renewable options (d= 0.3).

commercially though there are prototype PEM cells developed by SPIC Foundation in Chennai and BHEL. Even considering an optimistic estimate of $3000/kW for the fuel cell and the reformer, the cost of generation is still high. At a discount rate of 10% the PEM fuel cell competes with diesel engines at load factors of 70% and higher. At a 80% load factor the price of electricity from a PEM fuel cell is Rs. 4.7/kW h with the capital cost accounting for 70% of the cost of generation. In case of a discount rate of 30%, fuel cells do not compete at any load factor.

5. Cost of generation from renewables 5.1. Wind turbines Most of the installed wind capacity is grid-connected. The total installed capacity in September 2002 was 1702 MW (MNES, October 2002). Most of this (1639 MW) is from commercial projects. Individual machines range from 55 to 1250 kW. There are a number of Indian companies with foreign collaborators (Suzlon, Enercon, Vestas, REPL, BHEL) who are manufacturing and marketing wind turbines and generators. The wind resources of India have been mapped (data from 1000 monitoring stations throughout the country). A potential site is considered viable in case the average winds speeds at a height of 50 m is above 200 W=m2 . Wind speeds are high during the monsoon months (June to August) with relatively weak winds during the rest of the year. The viability of wind is critically dependent on the capacity factor that is site specic. The average capacity factor for wind installations in India can be computed by dividing the average power generation by the sum of the rated capacities of

ARTICLE IN PRESS
R. Banerjee / Energy Policy ] (]]]]) ]]] ]]] 5

Normalized Power Output (%)

all the installations. The annual generation in 20012002 was 1966 GW h resulting in an average capacity factor of only 13.3%. Table 2 shows the input data, the ALCC and costs of generation for wind. The average cost of generation (at the average load factor of 13.3%) is Rs. 5.14/kW h. This implies that several unviable wind turbines have been installed. This was due to the intial incentives based on capital subsidies and tax benets due to 100% depreciation. Incentives were not linked to generation. Prot making companies set up wind farms to avail of the tax benets. In many cases due to improper siting, the actual generation and capacity factors were low. There have been policy correlations. This resulted in a slow down of capacity additions during 19961998 followed by a more sustainable wind capacity addition. The initial experience had many unviable wind machines being installed in a hurry to avail tax benets without considering wind siting issues. Many of the machines were designed for European wind regimes that are different from the Indian wind regime (more seasonal and monsoon driven). The MNES has tried to improve the capacity utilization through technology development and emphasis on micro-siting. The MNES has established a dedicated research center for wind energy technology (CWET). India has a large wind resource assessment effort with more than 1000 wind monitoring stations. The wind energy programme operates commercially and is facilitated by the availability of innovative nancing schemes from the Indian Renewable Energy Development Agency (IREDA). In order to promote wind, the government has provided several incentives like 100% accelerated depreciation. Many state governments have provided capital subsidies (Andhra Pradesh, Maharashtra, Karnataka upto 20%), sales tax exemption. Most utilities permit wheeling, banking and buy-back (purchase price of Rs. 2.25/kW h in 19941995 with an escalation of 5% per year). Often wind farms are permitted to carry out third party sale. In states that have energy shortages, a company could install a wind farm to shield itself from mandatory power cuts. A chemical company in Gujarat (Excel Industries) invested in a wind farm in Dhag that had a low capacity factor of 10%. This was still considered to be a viable investment by the industry
Table 2 Cost of generation from wind 0.1 d 0:1 ALCC Rs. Rs./kW h d 0:3 ALCC Rs. Rs./kW h 5960 6.80 15 167 17.31 0.2 6048 3.45 15 255 8.71 0.3 6136 2.33 15 342 5.84 0.4 6223 1.78 15 430 4.40

since the Bhavnagar plant was exempt from power cuts by the Gujarat Electricity Board during periods of shortage because of its wind generation. A wind turbine is different from the non-renewable options for DG discussed earlier since its output uctuates during the day and over the year. Figs. 5a and b show the hourly variation and the monthly variation in the wind for a site on the west coast of India. This implies that wind needs to have a grid backup to meet the requirements for DG. The potential for wind power has been estimated to be 45,000 MW with 15,000 MW being the technical potential (assuming a low grid penetration). Though most state regulatory commissions are allowing a preferential tariff for wind power, the state electricity boards feel that large wind farms (e.g. more than 300 MW at Vankasuwde in Satara district of Maharashtra) supply the maximum output to the grid in the monsoon months when the system demand is at its lowest. An additional 5000 MW from wind is being targeted by 2012. For isolated systems wind diesel and WindDieselPV hybrids have a signicant potential. The wind energy programme in India has made the transition from demonstration to commercialization and can be further strengthened by indigenous technology development (especially for wind turbines of smaller rating in the kW range and controllers). It is expected that for the new installations optimal equipment selection and siting will result in higher capacity factors.

9 8
Avg wind speed m/s

7 6 5 4 3 2 1 0 0 4 8 12
Time of day (hours)

16

20

24

(a)
100 80 60 40 20 0

r be em ec er b D em ov N er ob ct ber O em pt Se t s gu Au ly Ju ne Ju ay M ril Ap ch ar M ary u br Fe ry a nu Ja

(b)

Month

Capital cost Rs. 50,000/kW, O&M cost Rs. 0.1/kW h, Life 20 years.

Fig. 5. (a) Daily variation of wind (Sanodar, West Coast); (b) Monthly variation of wind power (Sanodar, West coast India).

ARTICLE IN PRESS
6 R. Banerjee / Energy Policy ] (]]]]) ]]] ]]] Table 3 Cost of generation from solar PV Capital cost Rs./kW 200 000 23,930 13.66 24,039 10.98 250 000 29,803 17.01 29,912 13.66 300 000 35,676 20.36 35,785 16.34

Annual capacity factors of 38% have been reached at some of the Indian sites.

5.2. Solar photovoltaic The daily average solar insolation incident over India varies from 4 to 7 kW h=m2 depending on the location. Most regions in the country get about 300 clear sunny days a year. Figs. 6a and b show the variation in the solar insolation for a typical day and for different months during the year for Mumbai. A solar PV system converts the incident solar radiation directly into electricity using silicon based solar cells. For the modules available, the efciencies range between 1015%. In PV systems the capacity factor is decided by the insolation characteristics at the site with a maximum capacity factor of 25%. The advantage of PV is ease of operation and negligible operating cost. India has both monocrystalline silicon and polycrystalline silicon cells. Manufacturers include Tata-BP, Shell, BHEL, and Central Electronics limited. The total installed capacity of solar PV in India was 65 MW in 2002. This includes home lighting, street lighting, water pumping and stand-alone power systems. The grid connected systems account for only about 2.5 MW (31 systems average about 80 kW and largest about 240 kW peak). The annual production of PV cells in 19992000 was 9.6 MW and 11 MW of PV modules (Annual Reports MNES, 2000, 2001, 2002).
Solar radiation kW/m2 1.2 1 0.8 0.6 0.4 0.2 0 0 2 4 6 8 10 12 14 16 18 20 22 24 Time in hrs. 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 10 11 12 Month

LF 0:2 ALCC Rs. Rs./kW h LF 0:25 ALCC Rs. Rs./kW h

O&M Rs. 0.25/kW h, Life 20 years, discount rate = 10%.

The economics is computed for a grid connected system with no requirement for storage. For isolated systems there is an additional cost of storage batteries. Table 3 shows the cost of generation from solar PV. PV is expected to have niche markets in remote areas, islands etc. The main advantage is the maintenance free operation. For the PV systems installed for village electrication (Sunderbans in West Bengal) almost the entire capital cost has been provided as a capital subsidy. If subsidies are continued, there may be a number of remote villages electried through PV since systems are modular and can be quickly installed. However the costs are signicantly higher than the other renewable options. 5.3. Biomass gasiers operating gas engines Fuelwood, agricultural residues (rice husk, sugarcane trash, coconut shells...) and animal waste are the main biomass fuels available in India. The advantage of biomass fuels is that they are available throughout the country. Different biomass sources are available in different regions. Biomass (fuelwood, crop residues and cattle dung) accounts for about 40% of Indias primary energy use (TERI, 2000). At present biomass is mainly used for cooking in chulhas (cookstoves) with poor efciency. Aggregate estimates of biomass availability can be made from the crop production data and the residue to product ratio. Using this approach, Mukunda (1999) estimated the biomass produced in India in 19971998 to be 545 million tons. Of this, about 150 million tons is expected to be available for power generation. This is estimated to be made up of 23% rice straw, 18% wheat straw, 16% other straw, 15% bagasse and 12% plant stalks. These residues are estimated to have a generation capacity of 16,000 to 18,000 MW with a plant load factor of 68.5% (6000 h per year). MNES estimates a potential of 3500 MW from Bagasse based power generation and an additional 16,000 MW from other biomass that is already available (Annual Report MNES, 2002). In addition to residues that are available, it is possible to have dedicated plantations on waste land or degraded lands that are not normally used for agriculture. In social forestry programmes sustainable yields of

(a)
Avg Daily Solar Radiation (kWh/m2/day)

(b)

Fig. 6. (a) Daily solar radiation (Mumbai, May); (b) Variation in monthly solar radiation (Mumbai).

ARTICLE IN PRESS
R. Banerjee / Energy Policy ] (]]]]) ]]] ]]] 7

78 tons/hectare/year have been achieved (Ravindranath and Hall, 1995). The waste land available in the country has been estimated to be between 66130 million hectares (Mukunda, 1999). Using an average productivity of 5 tons/hectare/year and 100 million hectares of waste land the total biomass available annually is 500 million tons which can fuel power generation of 60,000 MW at a plant load factor of 68.5% (6000 h/ year). The options for conversion of biomass into electricity are combustion, gasication, IGCC, pyrolysis (Ganesh and Banerjee, 2001). The biomass can be converted into producer gas by gasication (partial combustion). Thermochemical gasication involves burning the biomass with insufcient air so that complete combustion does not occur and producer gas is formed. Producer gas is a mixture of carbon monoxide and hydrogen. Gasiers are classied as updraft or downdraft depending on the direction of ow of the biomass and the producer gas. In a downdraft gasier the biomass and the gases ow in the same direction (downwards). In a typical downdraft gasier the biomass is fed from the top. It passes through the gasier and undergoes the following sequence of processesdrying, pyrolysis, oxidation and reduction (Parikh, 1984). The gas formed is passed through a cooling and cleaning sub-system that usually consists of a cyclone for particulate removal and a scrubber for cooling and cleaning the gas (removing the tar). Some ash is formed from the oxidation reactions. The ash moves through the reduction zone and gets removed from the ash disposal system (grate and ash collection system). The typical composition of producer gas is 2022% CO, 1518% H2 , 24% CH4 , 911% CO2 and 5053% N2 (by volume). This is a low caloric value fuel with a caloric value of 10001200 kcal=Nm3 . India has signicant experience in atmospheric xed bed gasiers. About 1700 gasiers have been installed with a total installed capacity of 34 MW. The average gasier size is 20 kW. Biomass gasiers were initially developed for diesel replacement in agricultural pumpsets. Gasier models were indigenously developed around 1986. During the initial years of the National Demonstration Programme (19861994) the emphasis was on agricultural pumpsets of 5 and 10 hp rating. A feature of this programme was heavy subsidies on gasiers, pump-sets and diesel engines. It is estimated that the majority of installations (80%) during this phase become inoperative within one to three years of the system installation (ASCENT, 1998). Target beneciaries took little interest in the programme and reverted back to full diesel operation, after the initial few hundred hours of operation. Subsidies were misused to obtain a diesel engine pump-set at a cheaper rate. Despite this, there was important technology demonstration experience obtained in the installations that

continued to operate with gasiers. Since 1994 subsidies were reduced and were only available for the gasier. This initially resulted in a drop in the number of annual installations, but the programme is now more market oriented. Biomass gasiers have been developed either for wood or for rice-husk. Other fuels that have been used are cotton stalks, coconut shells, saw dust, palm shells, corn cobs. Installations range from 3 to 500 kW capacity. The biomass input required ranges from 5 to 500 kg/h for electrical outputs ranging from 5 to 500 kW. The largest installation size is 500 kW in Gujarat that is being connected to the grid. A 500 kW 5 100 kW rural electrication system has been installed at Gosaba in Sunderbans (West Bengal). A 100 kW rice husk based gasier has been installed in a rice mill in Andhra Pradesh. Almost all gasier systems installed are stand alone. Most installations use diesel engines in the dual fuel mode. There are a number of manufacturers Ankur Scientic (Ascent, Baroda), Netpro, Cosmos (Raipur), AEW and Tanaku. Decentralized Energy Systems India (DESI Power, 2003) has set up six projects as independent rural power producers (IRPP) in various parts of the country. The rst installation was at Orchcha in Madhya Pradesh (100 kW rating2 units of 50 kW each). DESI power estimates that a 100 kW IRPP will directly employ 11 persons and another 56 downstream jobs in new small scale industries (because of the availability of electricity). Instead of a diesel engine being operated in the dual-fuel mode that has a high operating cost and emissions because of the diesel fuel, it is preferable to opt for a dedicated spark ignition engine operated on producer gas. There are a few installations in the countryIISc Bangalore has set up a 100 kW dedicated engine in a milk chilling plant in Arnekal near Bangalore. Ankur has set up 100 kW gas engines in an industry near Baroda. Table 4 shows the input data for the calculation. Fig. 7 shows the cost of generation from this option. It is clear that this is preferable to diesel engines at load factors of 20% or higher. In case biomass is available, this can operate like a dispatchable power plant. Biomass gasier-engines appear to be a potentially cost effective DG solution. This has a potential for widespread diffusion since biomass can be obtained from waste lands using dedicated plantations. It is
Table 4 Input data for the biomass gasier-engine Gasier Capital cost (Rs./kW) Life Efciency 20 000 10 years 70% Engine-generator 33 000 20 years 35%

Biomass NCV = 3400 kcal/kg; Price Rs. 1/kg; Discount rate = 10%; O&M cost = Rs. 0.5/kW h.

ARTICLE IN PRESS
8 R. Banerjee / Energy Policy ] (]]]]) ]]] ]]]

12
Gas Engine Bioengine Diesel

National Bio Power Corporation (modelled on the lines of the National Thermal Power Corporation). 5.4. Bagasse cogeneration in sugar factories All sugar factories use bagasse as a fuel in their boilers to generate process steam and also to generate the electricity and shaft work required by the plant. Cogeneration is the simultaneous generation of power (electricity or motive power/shaft work) and process heat (steam). The process steam required in sugar factories is at low pressuresmost of the steam is required at 2 atm absolute (ata,) a small portion is required at 6 ata. Traditionally sugar factories have been designed to meet most of their power requirements during the crushing season from the bagasse itself. All sugar factories already have cogeneration of steam and power. However the steam generation pressures are low (usually 21 atm absolute (ata)). The mill turbines and power turbines are old and inefcient. If the steam generation pressures are increased by using a high pressure boiler, the sugar factories can export surplus power to the grid. A large number of options are possible. The options proposed have been to replace the milling turbines by efcient electric motors and the power turbine by an efcient backpressure turbine, increasing the generation pressure, using a condensing extraction turbine. An improved conguration that uses steam at 65 ata and passes through a condensing extraction turbine provides surplus power for export of 9.5 MW for a 2500 tons of cane crushed per day (tcd) plant (Smouse et al., 1998). For a given conguration, it is possible to select an optimal steam generation pressure (Raghu Ram and Banerjee, 2003). MNES estimates indicate a potential of 3500 MW net (additional) exportable capacity from the Indian sugar factories. Thirty four bagassebased cogeneration projects aggregating 210 MW have been commissioned till March 2001. The average export capacity of these plants is 6 MW per plant. The projects implemented have been with steam conditions of 60 ata (some are lower). New cogeneration projects designed with steam conditions of 87 ata and 515  C are being implemented (Annual report MNES, 2002. Bagasse based cogeneration has the problem that the mill operates only during the crushing season, 79 months a year.) Bagasse can be supplemented using other biomass fuels such as cane trash and rice hulls so that there is power export throughout the year. There are about 430 sugar mills in India. In terms of equivalent 2500 tpd mills, about 360 were in operation during 19961997 (Smouse et al., 1998). Based on the option discussed earlier, this would result in an exportable power output of about 3500 MW. Sugarcane production has been increasing at 35% per year. Hence the cogeneration potential can be expected to increase at

10 Cost of Generation Rs/kWh

0 0 0.1 0.2 0.3 0.4 0.5 0.6 Load Factor 0.7 0.8 0.9 1

Fig. 7. Cost of generation from bio-engine.

necessary to have assured biomass supply as the viability would depend locally on the biomass price. It is expected that the optimal ratings would be between 50500 kW. It is seen that the dedicated biomass red engine has a cost of generation comparable to the gas engine. The main advantage for this option is that it would operate on locally available resources, unlike the gas engine that would require natural gas transport and supply to the rural areas (this might increase the cost of energy from the gas engine). The viability of the biomass option is critically dependent on the availability and price of biomass. (The present value is Rs. 0.07/MJ or half the price of natural gas on an equivalent energy basis.) The main usage of biomass in the rural areas of India is for cooking. This is often collected from local wood lots or from areas near forests. There is no well developed market for biomass in most rural areas. Hence a dedicated biomass based power plant should ensure that it has a dedicated plantation attached to it. In case of isolated gasierengine systems it is essential that the system is coupled with an industrial load (cold storage, rice mill, oil mill etc.) so that the demand load factor can be improved and the revenue can be ensured. The operation of the gasier requires operator training. The institutional mechanism for cost recovery and plant operation needs a number of policy experiments. Most of the installations have been subsidized and operated by the technology supplier. Independent assessment of actual costs incurred and operating experiences need to be documented and disseminated before launching a large biomass gasier engine programme. The present manufacturer base and number of energy service companies (ESCOs) is sub-critical for a large scale programme. One possible solution is a setting up of a public-sector (on joint sector) national company the

ARTICLE IN PRESS
R. Banerjee / Energy Policy ] (]]]]) ]]] ]]] 9

this rate. The capital cost of the plant is around $450/ kW of output or $650/kW of exportable surplus. This works out to around Rs. 30,000 per kW of surplus power which is cheaper than setting up a new fossil fuel power plant. MNES provides an interest subsidy on term loans for cogeneration in sugar. The interest rate reduction is 3% for 80 ata and above, 2% for 6080 ata and 1% for 4060 ata. Most state governments have announced policies that x buy back rates, permit wheeling and banking Tata Energy Data Directory (1999). To encourage co-generation in co-operative and public sector mills a joint venture/ Independent Power Producer model has been proposed by MNES in the major sugar producing states. Capital subsidies are available for the rst projects of these types in each state. Many of the sugar factories are in the co-operative sector. They are traditional industries with limited technical capability. Often the sugar factory does not have the condence to operate high pressure steam based power plants. The capital investments are signicant, about Rs. 300 million for a 2500 tcd plant, and the sugar factories are hesitant to make these investments. At present boilers and turbines are available in India from suppliers like BHEL, Thermax, ISGEC, Triveni. Grid interconnection and recovery of dues from the State Electricity Board is perceived as a problem. Some states have announced a special tariff for Biomass Cogeneration (e.g the Maharashtra Electricity Regulatory Commissions recent tariff order). This tariff makes it viable for the sugar factory, even if it operates during the crushing season only. The bagasse based cogeneration option is only viable for large plants 2500 tcd or higher. Hence this option is suitable for 510 MW or more. This is the cheapest of the options considered, since only the incremental cost is charged to power generation. The efforts to promote bagasse based cogeneration seem to be providing the desired results. Access to soft loans for the capital investment and the development of a number of energy service companies (ESCOs) that could prepare detailed project reports and build, own and operate the plants may accelerate the installation of Bagasse based cogeneration (Table 5).
Table 5 Data and calculations for Bagasse cogeneration Incremental capital cost (Rs./kW) Life Boiler efciency Load factor Rs./kW h 30 000 20 years 70% 0.5 2.40

6. Conclusions The summary of the different options evaluated is presented in Table 6. The cost of generation of different DG options depends on the load factor. For some of the renewable options the system load factor is constrained by the supply availability. Among the non-renewable DG options considered, diesel engines are prevalent in India. This is because of the scarcity of capital and low load factors (use as backup power). In view of the government liquid fuel policy gas engines are likely to be the preferred option for DG. Gas engines are cost competitive in view of the relatively low natural gas price. These are likely to be the preferred option for DG in areas where natural gas is available. The existing engine manufacturers need to promote their gas engines in India. For PEM fuel cells and micro-turbines based on natural gas, there is a need to have a few demonstration projects and obtain experience with these technologies. Technology development and cost reductions could make either of these technologies cost-competitive. Among the renewable technologies considered wind energy is growing signicantly because of the supportive policy environment. For sites where the capacity factor is 30% or more, wind is competitive at present prices. Even though the comparison shows a price of Rs. 5.84/ kW h, the accelerated depreciation and tax benets provided make it a viable investment even at a selling price of Rs. 3/kW h. Biomass gasiers operating dedicated gas engines is a DG option that is almost cost effective and seems suited for rural areas. At present engine availability is a constraint. Engine manufacturers are not keen to develop producer gas engines as they are unsure of the volumes. Biomass availability, system standardization and institutional issues need to be addressed before this option can achieve widespread diffusion. A national level Bio-power corporation to provide technology solutions and operation and maintenance support may help this option reach its potential. Bagasse based cogeneration is cost-effective at present prices and is likely to provide about 3000 MW of surplus power to the grid. Solar PV does not seem to be a viable option for grid connected systems, at present prices. However, the technology is mature and requires low maintenance. This is the preferred option for small remote systems. For isolated systems hybrid systems of PV-Wind diesel are likely to be cost-effective (IIT Bombay, 2002). Accordingly to estimates of the Ministry of Power (MOP, 2003a, b) there are about 18,000 villages that are remote and difcult to connect to the grid. These villages can be electried by DG systems. This would result in a potential of about 500 MW of small isolated

0.4 2.60

0.6 2.27

Bagasse NCV = 3400 kcal/kg (dry basis); Price Rs. 1.50/kg; Discount rate = 10%; O&M cost = Rs. 0.5/kW h; 2500 tcd plant 9.5 MW export; 0.93 kg extra/kW h.

ARTICLE IN PRESS
10 Table 6 Summary of DG options Type Technology status C, I C D D C, I C, I C Gasier-I C, I Capacity factor Cost of generated electricity d 0:1 LF 0:5 Rs: 5:10=kW h LF 0:8 Rs: 4:85=kWh LF 0:5 Rs: 2:62=kW h LF LF LF LF 0:8 0:5 0:8 0:5 Rs: Rs: Rs: Rs: 2:29=kW h 3:24=kW h 2:82=kW h 6:64=kW h Comments R. Banerjee / Energy Policy ] (]]]]) ]]] ]]]

Diesel Gas engine Micro turbine fuelled by natural gas Fuel cell fuelled by natural gas Wind turbines PV BiomassGasier GasEngine Biomass Cogen

NR NR NR NR R R R R

N N N N 13% Avg Max 3038% Max 25% N 50% Higher if aux fuel is used

Existing base of more than 10,000 MW as backup. Relative price of Natural gas low Technology not proven in India Technology demonstration required 2000 MW already installed Niche applications Grid connected systems 2.5 MW installed (80240 kW)

LF 0:8 Rs: 4:68=kW h LF 0:2 Rs: 8:71=kW h LF 0:3 Rs: 5:84=kW h LF 0.25 Rs.17/kW h LF 0:5 Rs: 3:16=kW h LF 0:8 Rs: 2:59=kW h LF 0:5 2:40=kW h LF 0:6 2:27=kW h

About 300 MW installed export capacity in 2002

NRNon Renewable; IIndigenous; RRenewable; DDemonstration; CCommercially available technology; NNot constrained by the supply.

systems. The difculty for these systems is the need to match the supply with the load prole. For this constraint hybrids of two renewables or renewable-fossil can be shown to perform better than power plants based on a single technology. The isolated systems are likely to be costlier than the grid connected systems. In isolated systems, the DG system viability increases by the inclusion of an industry load (cold storage, rice mill...). Different institutional models adopted for DG in India have been reviewed in the Gokak Committee report (Gokak, 2003). The Sunderbans model involves a village committee that manages the project and collects bills from members, the local enterprise that operates and maintains the plant and the nodal agency (West Bengal Renewable Energy Development Agency). The Uttam Urja project in Rajasthan is an example of private Energy Supply Companies operating a DG project in collaboration with TERI and manufacturers. A large number of DG projects have to be initiated through different institutional mechanisms and the results tracked. The national strategy should involve demonstrations and pilot projects with some of the new technologies (PEM fuel cells, micro turbines), dissemination of successful implementation mechanisms, tracking of actual costs of generation from different DG options and promotion of the cost-effective options. This strategy could result in increasing the share of DG to 10% of the total electricity by 2012, as envisaged by the Government.

Acknowledgements The author is grateful for nancial support from the Carnegie Mellon Electricity Industry Center for the duration of this work.

References
Ackerman, T., Andersson, G., Sodder, L., 2001. Distributed generation: a denition. Electric Power Systems Research 57, 195204. ASCENT, 1998. Status of Biomass Gasication Technology, India, October 1998; http://www.bgtechnologies.net/ankur.htm. Borbely, A.M., Kreider, J.F., 2001. Distributed Generation, The Power Paradigm for the New Millennium. CRC Press, Boca Raton, USA. DESI Power, 2003. www.desipower.com. Ganesh, A., Banerjee, R., 2001. Biomass pyrolysis for power generationa potential technology. Renewable Energy 22, 914. Gokak Committee Report on Distributed Generation, 2003. Submitted to the Ministry of Power, Government of India, 2003; available at powermin.nic.in. ICRA, 2003. Indian Credit Rating Agency New Delhi, Rating Feature, The Indian Natural Gas Industry, January; available on the web at //www.icraindia.com/biz-arch/naturalgasindustry-changes.pdf. Ministry of Non Conventional Energy Sources, 2001. Renewable Energy in India, Business Opportunities, Government of India, March 2001. Ministry of Non Conventional Energy Sources, 2002. Wind power development in India: Towards global leadership; New Delhi, October 2002. Ministry of Non Conventional Energy Sources, Annual Reports, New Delhi, 1993, 2000, 2001, 2002. Ministry of Power, 2001. Blueprint for Power Sector Development, Government of India, New Delhi; available at powermin.nic.in.

ARTICLE IN PRESS
R. Banerjee / Energy Policy ] (]]]]) ]]] ]]] Ministry of Power, 2003a. Annual Report 20022003, Government of India, New Delhi. Ministry of Power, 2003b. Discussion Paper on Rural Electrication Policies, November 2003, Government of India, New Delhi. Mukunda, H. S., 1999. A biomass option for enhancing energy security. NIAS Report R4-99, National Institute of Advanced Studies, IISc Bangalore. Parikh, P.P., 1984. State of the Art Report on Gasication of Biomass, Report submitted to DNES, Government of India. Raghu Ram, J., Banerjee, R., 2003. Energy and cogeneration targeting for a sugar factory. Applied Thermal Engineering 23, 15671575. Ravindranath, N.H., Hall, D.O., 1995. Biomass Energy, and Environment, A Developing Country Perspective From India. Oxford University Press, New York. Resource Dynamics Corporation, 2001. Assessment of Distributed Generation Technology Applications, Prepared for Maine Public 11 Utilities Commission by Resource Dynamics Corporation, Vienna, VA, USA; available on the web at //www.distributed-generation.com. Smouse, S.M., Staats, G.E., Rao, S.N., Golman, R., Hess, D., 1998. Promotion of biomass cogeneration with power export in the Indian sugar industry. Fuel Processing Technology 54, 227247. Tata Energy Data Directory, 1999. Tata Energy Research Institute, New Delhi, 2000. USDOE, Department of Energy, 1998. Fuel Cell Handbook, Fourth Edition, November 1998, Ofce of Fossil Energy, Federal Energy Technology Centre, Pittsburgh, USA; available on the web at // www.fetc.doe.gov/publications. USDOE, 2003. Energy Prices from the Energy Information Administration, US Department of Energy; available on the web at http:// www.eia.doe.gov/price.html.