Optimization of decentralized energy systems using biomass resources for rural electrification in developing countries

by

Diego Silva Herran, Toshihiko Nakata Department of Management Science and Technology, Graduate School of Engineering Tohoku University Aoba-yama 6-6-11-815, Sendai, 980-8579 Japan E-mail: [email protected] Phone/Fax : +81-22-7956987

Abstract This research presents an optimization model describing decentralized energy systems for rural electrification based on local biomass resources, looking forward the improvement of electricity access conditions in rural areas in developing countries. Access to modern energy in rural areas is a basic component in order to improve living conditions of poor people in developing countries and achieve the Millennium Development Goals. The contribution of energy to rural development in developing countries has been scarcely addressed by studies on energy systems using energy models based on bottom-up approaches. Current energy models primarily focus on system’s technical and economic performance, as well as system’s potential to mitigate carbon emissions, giving less importance to income and energy use differences between rural and urban areas. The purpose of this research looks at the design of a linear programming (LP) optimization model for rural electrification with decentralized energy systems based on local biomass resources, integrating rural/urban differences into the analysis. The performance of the system utilizing agricultural wastes and forest biomass is studied as an alternative to grid extension and utilization of diesel fuel in diesel generators for the areas not interconnected to the grid. The target area corresponds to a region in Colombia, South America, partially covered by the electricity grid and where over 15% of the population has no access to electricity. Inclusion of biomass-based electricity generation into the energy system increased significantly the electricity costs, which rose from 7.4 cents/kWh in the baseline case, to 10.4 cents/kWh in a case considering electricity supply entirely based on local biomass resources. In the same way, the ratio of overrun costs to total costs of the energy system increased from 21% to 44%. On the other hand, greater dependence on local biomass for electricity 1

supply may grant a maximum CO2 emissions reduction of over 120,000 tons per year, and a smaller disparity in the conditions of electricity supply between socio-economic groups in the household sector. The biomass-based energy system may result in other benefits, such as the utilization of agricultural wastes and the promotion of productive activities, although they were not quantified in this research.

1. Introduction Access to electricity in developing countries is an important driver of rural development given that access to more convenient forms of energy contributes to the improvement of quality of life among poor population. Rural electrification is also important to reduce the considerable differences in living conditions existing between rural and urban areas of developing countries. In fact, access to electricity and other modern forms of energy like liquefied petroleum gas (LPG) are correlated to the achievement of the Millennium Development Goals (MDGs), a set of global targets promoted by the United Nations looking forward the reduction of poverty and better living conditions for the poor (Modi et al. 2005). Currently over 1.6 billion people have no access to electricity, the large majority of them living in rural areas of developing countries, and depend mainly on traditional fuels to cover their energy needs (IEA 2006). Electrification in rural areas of developing countries, and in particular in the case of remote areas, is difficult due to low population densities, highly dispersed location of populated centers, low energy consumption levels per capita and poor road infrastructure which constrains transportation (World Bank 1996). This makes conventional rural electrification programs based on extension of the electricity grid and decentralized schemes with diesel generators expensive or even economically not feasible, requiring direct intervention of government generally in the form of subsidies. In rural areas where energy resources are widely available in the form of agricultural wastes and forest biomass, decentralized electrification using local resources is a more suitable alternative since it avoids the necessity of extending transmission lines to dispersed populated centers, reduces the road transportation limitations within these areas, and promotes local development through the introduction of the production chain of biomass energy. Moreover, biomass for electricity generation, in contrast to diesel generation and other fossil fuel based electricity supply methods, does not contribute to climate change as it is considered neutral source of carbon emissions. Although electrification based on biomass energy has lower conversion efficiencies and higher investment costs compared to conventional electricity generation schemes, and its application to low income areas is limited in terms of financial sustainability, decentralized electrification based on biomass resources may have 2

a more significant impact on rural development due to the promotion of new income opportunities around the use of local resources. There are several examples of biomass applications for decentralized electrification in developing countries, especially in Asia and some Latin American countries. Small scale applications for rural electrification using biomass gasification processes exist in Thailand, India, China and Brazil, among others (Knoef 2005; Sims 2002; Silveira 2005). For the evaluation of decentralized electrification two aspects of energy systems in developing countries are of special interest, namely the extension of access to electricity and the large differences between urban and rural areas. The urban/rural divide can be interpreted beyond the geographical location of rural and urban areas, including the differences in income levels. Distinguishing between urban and rural areas helps to describe the income opportunities according to the availability of biomass energy resources. Incorporating income differences may allow the characterization of the energy system’s performance based on the impact of electricity costs on energy expenditure in households with different incomes. They are adopted in this study as indicators of rural development in developing countries. In this research a linear programming (LP) model is introduced in order to evaluate a decentralized energy system for rural electrification in developing countries using local biomass resources, incorporating differences in energy consumption and income levels between urban and rural areas. The energy system considered comprises the utilization of agricultural wastes and forest wastes for electricity generation by means of several energy conversion technologies using biomass (direct combustion of biomass in boilers, gas turbines coupled with gasification unit and diesel engines coupled with a pyrolysis unit). The target area is a region in Colombia, South America, characterized by abundant biomass resources in the form of agricultural wastes, with partial coverage of the electricity grid, where over 15% of population has no access to electricity. Currently, electricity from the grid covers the interconnected area, and diesel generators supply the non interconnected area.

2. Models for decentralized electrification in rural areas of developing countries In order to evaluate the performance and the impact of decentralized energy systems for rural electrification using biomass resources in developing countries, it is necessary to take into account certain characteristics of energy systems in these countries. Many of these features, presented in Table 1, have been commonly overlooked in the literature dealing with energy planning supported by mathematical models (Urban, Benders, and Moll 2007). It is worth noting that among these factors off-grid renewable energy, rural energy programs and energy issues related to poverty 3

Table 1 Features of energy systems in developing countries Feature Performance of power sector Supply shortages

Examples Sub-optimal system configuration due to plant breakdown, outages and voltage fluctuations. Poor condition of equipment, inadequate operational and maintenance performance, organizational and technical problems.

Electrification

Lack of electricity access in many areas

Traditional bio-fuels

Traditional biomass (firewood, dung, agricultural wastes) are predominant fuels.

Urban-rural divide/ urbanization

Drastic differences in rural and urban areas (services and infrastructure).

Informal economy

Non-monetary transactions, illegal activities, tax evasion.

Structural economic change

Rapid shift from agricultural to services economy

Investment decisions

Inadequate planning techniques.

Subsidies

Abuse or inadequate use of subsidies.

Adapted from Urban et al. (2007).

aspects are rare among applications (van Ruijven et al. 2008). Most models describing decentralized energy systems are designed in order to find the optimal mix of energy resources and technologies under a certain objective function and set of constraints. For example, minimum system costs and minimum level of emissions stemming from system operation are common objective functions. The analytical approach generally used in these models is single-period optimization (Ashok 2007; Karki, Mann, and Salehfar 2008, 2006). In addition to optimization, there are studies deploying simulation and geographic information systems (GIS) methodologies that give more emphasis to supply stability and optimal allocation of resources (Yue and Wang 2006; Underwood et al. 2007; Nfah, Ngundam, and Tchinda 2007). Decentralized energy systems have also been analyzed by means of multi-criteria and multi-objective methodologies (Silva and Nakata 2008; Cherni et al. 2007; Kanniappan and Ramachandran 2000; Chetty and Subramanian 1988; Hiremath, Shikha, and Ravindranath 2007; Pohekar and Ramachandran 2004). Large availability of biomass resources in rural areas makes the efficient allocation of energy resources an important aspect of energy modelling in developing countries. The scope of the analytical procedure may vary from taking into account only the efficient distribution of energy resources to cover specific energy needs, to including land-use patterns, consumption rates of local agricultural and livestock-derived resources among others (Rubab and Kandpal 1996; Hartter and 4

Boston 2007; Painuly, Rao, and Parikh 1995; Parikh 1985; Parikh and Ramanathan 1999; Thankappan, Midmore, and Jenkins 2006).

3. Construction of a rural energy model In this study an optimization model is constructed for the analysis of decentralized energy systems for rural electrification in developing countries, making emphasis on the utilization of locally available biomass energy resources in a region of Colombia. The contrasting conditions of rural and urban areas in the target area with respect to energy consumption are introduced in the model, describing several groups of electricity consumers in the residential sector as socioeconomic strata which indicate different income levels.

3.1. Target area The target area is a region in Colombia called Meta department. This region is a suitable case for analyzing the potential of decentralized electrification to narrow the gap between rural and urban areas in developing countries. Compared to other regions of Colombia, Meta has a considerable area outside the interconnected system in spite of its proximity to the capital city of the country, as showed in Figure 1. Moreover, agricultural activities in large plantations in the region

Figure 1 Map of the target area of the study

5

Table 2 Main characteristics of the target area NIS

NIZ

Total

Population

724,929

18,668

743,597

Electricity consumption (MWh/yr)

182,174

2,473

184,647

86

70

85.9

0.070

0.114

n.a

Electrification rate (%) Electricity price (US$/kWh) Calculated based on data from UPME (2000).

generate considerable quantities of agricultural wastes, which offer an opportunity to produce electricity from local resources. This region is partially served by the national electricity network, called national interconnected system (NIS). Over two thirds of the electricity generated in the interconnected system is based on large scale hydropower plants. Over two thirds of the electricity generated in the interconnected system is based on large scale hydropower plants. The rest of the electricity is supplied mainly by coal power plants (UPME 2008). Remote areas which are outside the interconnected area, referred to as non interconnected zones (NIZ), depend on a low quality electricity service based on diesel generators using fossil fuels, with only eight hours of average daily operation. Dispersed distribution of population and poor road infrastructure limits the feasibility of electricity grid extension and the supply of fuels within the NIZ. There are 170,577 people without access to electricity in the region. Table 2 gives general information of the target area (UPME 2000; Ramirez Gomez 2007).

3.2. Energy system considering rural-urban differences The energy system considered for rural electrification in the study, illustrated in Figure 2, describes the demand sectors for rural and urban areas according to their location within or outside the interconnected electricity grid. Households are divided into socio-economic strata in concordance with the current categorization of households for setting electricity consumption tariffs in the target area. Households of upper strata, i.e. 5 or 6, correspond to higher income groups, while lower strata correspond to lower income groups. In addition to the residential sector, other sectors are considered in the energy system, including the industrial, commercial and official sectors. Three types of households with electricity are differentiated in the NIS. NIS-Urban corresponds to cities and large townships, NIS-Rural describes medium size townships, and NIS-Populated center are small populated centers. Within the NIZ only households in townships, labeled NIZ-Rural, are 6

Energy resources

Foreign resources Electric grid

Energy conversion technologies

Electricity transmission and distribution

Centralized electrification

Areas interconnected to electricity grid

Electricity demand

Urban areas

Fossil fuels

Decentralized electrification Local resources Agricultural wastes Forest biomass Other renewable

Rural areas

Diesel generation Areas outside electricity grid

Biomass technologies

Remote rural areas

Other renewable technologies

Figure 2 Schematic of the rural energy system

considered, excluding farms outside villages. Electricity consumption and electricity prices for each location and socio-economic stratum are listed in Table 3. In addition to diesel generation with diesel fuel, energy conversion technologies included in the model are boilers based on direct combustion of biomass (Direct Combustion), gasification unit coupled with a gas engine (Gasification) and diesel engine using fuels resulting from pyrolysis unit (Pyrolysis). All generation units are of 2MW scale operating at 50% of the total capacity. Economic evaluation assumes a discount rate of 10% and 20 years operating time. CO2 emissions factor derived from electricity from the grid and diesel generation are 0.439 kg/kWh and 0.882 kg/kWh respectively. Distribution and transmission costs for electricity are not considered. Main features of technologies are presented in Table 4 (Solantausta and Huotari 1999; UPME 2000). Energy resources are divided into two groups: foreign resources, which comprehend electricity from the grid and diesel fuel; and local resources, which include agricultural wastes such as rice husk, sugarcane plantations wastes and bagasse, and forest wastes from planted and natural forests. The quantities and prices of the energy resources are presented in Table 5 (UPME 2003). Although sugarcane agriculture and harvesting of forests occurs only at low scale in the region, soil and weather conditions offer good conditions for a large scale production. Therefore, estimates from a study on the future potential of biomass have been deployed. For example, the quantity of bagasse and wastes from sugarcane plantations correspond to 0.1% of the potential amount produced if all

7

Table 3 Electricity demand in the target area

Location

Electricity demand

Electricity price by sector/socio-economic

MWh/yr

stratum US¢/kWh

Residential

Other

Residential / Stratum

Other

1

2

3

4

5

6

NIS-Urban

160,028

31,983

3.5

4.2

6.0

7.0

8.4

8.4

8.4

NIS-Rural

22,146

6,205

3.5

4.2

6.0

7.0

8.4

8.4

8.4

0

5,984

3.5

-

-

-

-

-

8.4

2,473

1,044

6.0

-

-

-

-

-

14.4

184,647

45,216

-

-

-

-

-

-

NIS-Populated centre NIZ-Rural Demand-Total

-

Calculated from data in UPME (2000) and SSP-SUI database.

Table 4 Characteristics of technologies Conversion technology

Efficiency

Capital cost

O&M costs

%

US$/kW

US¢/kWh

Diesel generation

30

300

1.70

Direct Combustion

17.5

2,300

0.05

Gasification

23.9

4,200

0.07

Pyrolysis

24.7

3,600

0.21

Source: UPME (2000) and Solantausta and Huotari (1999)

Table 5 Availability and cost of energy resources Resource

Diesel fuel

Stock

Cost

Energy content

GJ/yr

US$/kg

MJ/kg

n.a.

0.290

45.5

Rice husk

609,376

0.080

13.9

Bagasse

129,663

0.010

8.9

Sugarcane wastes

346,833

0.014

15.0

Natural forest wastes

3,329,050

0.030

16.8

Planted forest wastes

2,331,371

0.030

16.8

Diesel fuel cost in the NIZ is US$0.53/kg. Calculated based on data from UPME (2003).

8

agricultural land in the target area were dedicated to sugarcane production.

3.3. Linear programming formulation Rural electrification programs in developing countries are likely to incur in total costs larger than the revenues from the consumption bills due to the low levels of income and electricity consumption per capita of rural households, and thus, involve overrun costs. Therefore the energy system is designed looking forward the minimization of the resulting total overrun costs, B, in order to provide electricity equitably to users in both the interconnected and the non interconnected areas. In addition, overrun costs are minimized considering income differences indicated by energy expenditures across several socio-economic strata. The following equations explain the objective function and the constraints of the linear programming formulation.

M in im ize B

(1)

B = ( T o ta l _ co s ts ) − ( T o ta l _ reven u e ) =

∑∑b k

⎡∑ ⎢ j Subject to ⎢ ⎢⎣

∑

∑c η ∑d ijk

j

i

kl

l

s jk =

j

∑

kl

(2)

l

q ijk ⎤ ⎥ ⎥ × d kl − bkl = p kl d kl ⎥⎦

d kl

(3)

(4)

l

s jk =

∑η

j

q ijk

(5)

i

∑∑q j

ijk

≤ ri

(6)

k

bkl : overrun costs in each sector/stratum (US$/yr) B : total overrun costs (US$/yr) cijk : unit electricity generation cost (US$/GJ) dkl : electricity demand in each sector (GJ/yr) pkl : unit price of electricity paid by each sector/stratum (US$/GJ) qijk : primary energy resource allocated to each conversion plant (GJ/yr) ri

: stock of energy resource (GJ/yr)

sjk : electricity supplied by each conversion technology (GJ/yr) 9

ηj : electricity conversion efficiency i

: energy resource

j

: energy conversion technology

k

: location (within the NIS or the NIZ and urban or rural)

l

: energy demand sector and socio-economic stratum

The constraints of the model include the balance between electricity demand and supply in each sector, dkl, the energy balance around electricity plants, the complete utilization of agricultural wastes, and the maximum resource utilization indicated by the stock of each energy resource, ri. An additional set of constraints represents the limitations of electricity supply options with respect to geographical location. For example, rice husk and bagasse for electricity generation purposes are only available in significant quantities in large scale rice mills and sugarcane processing factories respectively, located in areas within the NIS. The price of electricity is US$0.070/ kWh for the areas in the NIS and US$0.114/kWh for the areas in the NIZ. The linear programming model is solved using the General Algebraic Modeling System (GAMS) software.

3.4. Case settings Several cases are proposed in order to study the progressive penetration of decentralized electrification in the target area. a)

Baseline case: this case represents the current situation of the energy system in the target area, in which electricity is supplied by the electricity grid to areas within the NIS, and by diesel generators using diesel fuel to areas within the NIZ; it is assumed that households currently without electricity connection are supplied by new diesel generation systems.

b)

Biomass-Remote case: decentralized electrification using biomass is introduced for households without electricity in the NIZ, displacing the use of diesel generators.

c)

Biomass-Rural case: decentralized electrification with biomass provides electricity to all rural areas in both the NIS and the NIZ, in addition to newly electrified households.

d)

Biomass-All case: all electricity demand is supplied by means of decentralized electrification with biomass. This case is used to estimate the potential of biomass resources to satisfy electricity needs in the target area.

10

4. Results The cases analyzed were compared in order to get some understanding of the implications that introduction of decentralized electrification with local biomass resources may have on the design of an energy system for a region consisting of urban and rural areas of contrasting living conditions. Such contrast in quality of life is pictured by means of the convenience of the electricity supply in locations with and without connection to the electricity grid, and estimated with the electricity cost and the overrun costs of the energy system.

4.1. Energy system structure Energy resources allocation and the structure of the energy system for the different locations are presented in Figure 3 and Figure 4, respectively. The energy system for decentralized electrification with biomass resources was conditioned to the complete utilization of agricultural wastes. This condition gave lower preference to the utilization of forest wastes, which were part of the energy system only in the Biomass-All case, and to a lesser extent in the Biomass-Rural case. In the Biomass-Remote case, promotion of complete utilization of agricultural wastes lead to the displacement of electricity from the grid with direct combustion of rice husk in urban areas of the interconnected area. Although there were cheaper alternatives for electricity generation in these areas, utilization of rice husk instead of other agricultural wastes in other areas where this resource was available would have resulted in a more expensive electricity supply. Among energy conversion technologies for the utilization of biomass, only direct combustion in boilers formed part of the energy system. The gains in performance due to better energy conversion efficiency of alternatives based on gasification and pyrolysis conversion processes were not able to offset the elevated capital costs of these technologies. Improvements in conversion efficiency and costs according to the scale of electricity generation plants may provide a higher participation of gasification and pyrolysis conversion technologies in the energy system configuration.

4.2. System performance Given that, in general, financial sustainability of rural electrification is highly dependent on government subsidies, the optimization of the biomass-based energy system for decentralized electrification was performed considering the minimum overrun costs for the energy system in the target area. Figure 5 presents the electricity cost for the different cases. Total overrun costs are presented in Figure 6. Main system’s performance results are presented in Table 6. The analysis 11

Baseline

Biomass-Remote

12

Biomass-Rural

Figure 4 Electricity supply and conversion technologies by location

Biomass-All

NIZ-Non_elec

NIZ-Rural

600

NIS-Non_elec

Biomass-Rural

NIS-Pop_Center

NIS-Rural

NIS-Urban

NIZ-Non_elec

NIZ-Rural

NIS-Non_elec

Biomass-Remote

NIS-Pop_Center

NIS-Rural

NIS-Urban

NIZ-Non_elec

NIZ-Rural

NIS-Non_elec

Baseline

NIS-Pop_Center

NIS-Rural

NIS-Urban

NIZ-Non_elec

NIZ-Rural

NIZ-Non_elec

NIZ-Rural

NIS-Non_elec

NIS-Pop_Center

NIS-Rural

NIS-Urban

NIZ-Non_elec

NIZ-Rural

NIS-Non_elec

NIS-Pop_Center

NIS-Rural

NIS-Urban

NIZ-Non_elec

NIZ-Rural

NIS-Non_elec

NIS-Pop_Center

NIS-Rural

NIS-Urban

NIZ-Non_elec

NIZ-Rural

NIS-Non_elec

NIS-Pop_Center

60%

NIS-Non_elec

NIS-Rural

NIS-Urban

Resource allocation

80%

NIS-Pop_Center

NIS-Rural

NIS-Urban

Secondary energy (103 GJ/yr)

100% Planted_f orest_waste

Natural_f orest_waste

Sugarcane_waste

Bagasse

Rice_husk

Diesel_f uel

Elec_Grid

40%

20%

0%

Biomass-All

Figure 3 Allocation of energy resources by location

800 Pyrolysis

Gasif ication

Dir_Comb

Diesel_Gen

400 Elec_Grid

200

0

Unit electricity cost (US$/kWh)

0.20

0.15 NIS-Urban NIS-Rural NIS-Pop_Center

0.10

NIS-Non_elec NIZ-Rural NIZ-Non_elec

0.05

Average

0.00 Baseline

Biomass-Remote Biomass-Rural

Biomass-All

Figure 5 Energy system’s electricity cost

Overrun costs by location (10 3 US$/yr)

12,000 NIZ-Non_elec

10,000

NIZ-Rural NIS-Non_elec

8,000

NIS-Pop_Center NIS-Rural

6,000

NIS-Urban

4,000

2,000

0 Baseline

Biomass-Remote

Biomass-Rural

Biomass-All

Figure 6 Total overrun costs by case

showed that, even in the Baseline case, the electricity service involves overrun costs, and that only the cost of the electricity supplied by the grid is low enough to avoid the need for financial support. For the baseline case, average unit electricity cost was US$ 0.074/kWh. Overrun costs amounted nearly US$4 million per year, and were observed principally in urban areas and newly electrified households within the area interconnected to the grid. Over 80% of these overrun costs were associated to the two lowest socio-economic groups of these two groups of users. They 13

Table 6 System performance outcomes Resource

Baseline

BiomassRemote 6,539

BiomassRural 6,588

Biomass-All

Overrun costs (103 US$/yr)

3,889

Unit cost (US$/kWh)

0.074

0.084

0.085

0.104

21

31

32

44

117,452

85,211

84,331

0

0.00

0.13

0.13

0.48

Ratio of overrun costs to total costs (%) Total CO2 emissions (kg-CO2/yr) Emissions reduction (kg-CO2/kWh)

11,463

comprehend 40% and 47% of the electricity demanded by the population in the lowest socioeconomic levels, respectively. The progressive inclusion of biomass resources into the energy system configuration raised in a proportional manner the unbalance between costs and revenues for the electricity supply. Overrun costs occurred mostly in households which belong to the lower socio-economic stratum and to users in other sectors. Promotion of rural electrification with biomass in remote areas resulted in lower electricity costs than diesel fuel-based electricity generation in these areas. Extending biomass-based electrification to all rural areas resulted in no significant changes in the overrun costs of the energy system, as indicated by the outcomes of the Biomass- Rural case and the Biomass-Remote case. Nevertheless, reallocation of biomass energy sources to rural areas increased considerably the unit costs of electricity supply in areas outside urban centers within the interconnected area. If all the electricity demand in the target area were supplied by local biomass energy, overrun costs would rise to more than US$11 million per year. Average electricity generation costs in cases considering the use of biomass for electricity generation were between US$0.084/kWh andUS$0.104/kWh. Emission reductions achieved by a biomass based energy system for decentralized electrification, taking into account the emissions levels in the Baseline case, correspond to over 120,000 t-CO2 per year. The total electricity costs, the overrun costs, and the CO2 emission of the energy system are presented in Figure 7, showing the changes in these values with respect to the share of biomass energy resources in total electricity generation.

5. Discussion The wide availability of biomass resources in the target area of the study may satisfy all electricity needs, even if electricity from the grid were to be replaced by biomass based electricity. The potential of biomass-based electrification was highly constrained by the elevated costs of

14

Cost (US$/kWh)

100,000 20,000 75,000

50,000 10,000 25,000

CO2 emissions (t-CO2 / yr)

125,000

30,000

Electricity cost Overrun cost CO2 emissions

0

0 0

20

40

60

80

100

Share of biomass in electricity generation (%)

Figure 7 Sensitivity analysis results

electricity generation using biomass conversion technologies. Large part of the costs was comprised by the capital costs of these technologies. In addition, the lower electricity cost estimated for the baseline energy system, can be attributed to the low costs of electricity from the grid, the major electricity supply source in the Baseline case. Therefore it is important to evaluate financial mechanisms, like subsidies, in order to estimate the feasibility of the penetration of biomass energy in the target area. An alternative approach is to internalize the benefits granted by biomass based electrification in terms of reduction of greenhouse gases (GHGs) emissions. Conventional approaches based on electricity costs can only highlight the advantages of local biomass through the differences in fuel costs, overlooking the benefits granted to rural areas. Other important aspects that point out the suitability of decentralized electrification with biomass have not been included in the linear programming formulation used in the study. For instance, the fact that supply of diesel fuel is unreliable due to transportation problems was only indicated by the comparatively higher price of this fuel in the NIZ than in the other regions. Another benefit of local biomass energy neglected by the model was the improvement of local economy through the utilization of wastes in productive activities. Energy generation from agricultural wastes adds new income opportunities for the rural population and diminishes the environmental impact produced by inappropriate disposal of these materials in open fires or dumps.

15

Regarding the reduction of the gap in living conditions existent between rural and urban areas, the importance of decentralized electrification with local biomass resources can be outlined by analyzing the changes in electricity access conditions for households. Reduction in differences in improved access to electricity can be illustrated by the balance between the electricity supply costs and the need of external financial support, such as subsidies, across groups of population with different income levels. For this purpose, the ratio between the overrun costs and the real value of electricity consumption, equivalent to the cost of electricity supply in each sector and socioeconomic level, was used as an indicator of the financial sustainability of the energy system. For instance, a 50% ratio value for a given group of users would mean that a quantity of money 50% larger than the payments for electricity consumption is needed to cover the costs of the real service. Thus, the greater the value of the ratio, the greater the amount of financial support needed to provide access to electricity supply in accordance with the capacity of payment for a given group of consumers. A ratio equal to or lower than 0% would mean that payments can offset the costs of electricity consumed. Figure 8 and Figure 9 show the average ratios of overrun costs to total costs with respect to each location and each socio-economic stratum, respectively. As can be observed from the graphs, in the present condition of the energy system in the target area, indicated by the Baseline case, ratios’ values increase with respect to lower income in households of lower socio-economic levels. When comparing the ratios’ values for locations in the NIZ and in the NIS, it is found that the gap

Ratio of overrun costs to total costs (%)

100

75 NIS-Urban NIS-Rural

50

NIS-Pop_Center NIS-Non_elec NIZ-Rural

25

NIZ-Non_elec Average

0

-25 Baseline

Biomass-Remote

Biomass-Rural

Biomass-All

Figure 8 Ratio of overrun costs to total electricity costs by location 16

Ratio of overrun costs to total costs (%)

75

Resid-Strat1

50

Resid-Strat2 Resid-Strat3 Resid-Strat4

25

Resid-Strat5 Resid-Strat6 Average

0

-25 Baseline

Biomass-Remote

Biomass-Rural

Biomass-All

Figure 9 Ratio of overrun costs to total electricity costs by socio-economic strata

among the average ratio values tends to narrow with an increased use of biomass for electricity generation. This occurs at the expense of a slight increase in need of financial support due to the higher costs of the biomass-based electrification indicated by the LP formulation. This trend towards a more equitable condition in the access to electricity at higher electricity costs is outlined by the trend in the overall average ratio of overrun costs to total costs of the energy system.

6. Conclusion The estimated availability of biomass resources in the target area allows for the introduction of energy systems based on local resources. However, electricity costs are four to five times higher than with the baseline energy system. Analyzing the system’s economic performance based on the price instead of the electricity cost is necessary to obtain a clear image of the feasibility of biomassbased energy system in rural areas. Incorporation of differences among rural and urban areas into the energy model showed that increased shares of decentralized electrification with local biomass resources in the target area reduces the disparity in the electricity access conditions among geographic locations and socio-economic levels. However, achievement of this benefit results in higher overrun costs, principally caused by the higher costs of biomass energy conversion

17

technologies compared to conventional electrification schemes based on diesel generation and electricity from the grid. In order to arrive at a more comprehensive outlook of the role of decentralized electrification with local biomass resources in rural development it is necessary to include other objectives related with the improvement of energy system’s efficiency, emissions reduction potential, electricity plant scale considerations and geographical location of resources, among other factors. Future improvements regarding inclusion of energy distribution aspects of decentralized systems are also necessary, as well as the estimation of socio-economic impact in rural areas.

References Ashok, S. (2007). "Optimised model for community-based hybrid energy system", Renewable Energy, 32 (7):1155-1164. Cherni, J.A. I. Dyner F. Henao P. Jaramillo R. Smith and R.O. Font. (2007). "Energy supply for sustainable rural livelihoods. A multi-criteria decision-support system", Energy Policy, 35 (3):1493-1504. Chetty, K.M. and D.K. Subramanian. (1988). "Rural energy consumption patterns with multiple objectives", International Journal of Energy Research, 12:561-567. Hartter, J. and K. Boston. (2007). "An integrated approach to modeling resource utilization for rural communities in developing countries", Journal of Environmental Management, 85 (1):7892. Hiremath, R.B. S. Shikha and N.H. Ravindranath. (2007). "Decentralized energy planning; modeling and application - a review", Renewable & Sustainable Energy Reviews, 11 (5):729-752. IEA. (2006). World energy outlook 2006. Paris: OECD/IEA. Kanniappan, P. and T. Ramachandran. (2000). "Goal programming model for sustainable electricity production from biomass", International Journal of Energy Research, 24 (1):1-18. Karki, S. M.D. Mann and H. Salehfar. (2006). "Substitution and Price Effects of Carbon Tax on CO2 Emissions Reduction from Distributed Energy Sources" (paper presented at Power Systems Conference: Advanced Metering, Protection, Control, Communication, and Distributed Resources, 2006. PS '06, Location). ———. (2008). "Environmental implications of renewable distributed generation technologies in rural electrification", Energy Sources Part B-Economics Planning and Policy, 3 (2):186195. Knoef, H.A.M. (2005). Handbook of biomass gasification. The Netherlands: Biomass Technology Group. Modi, V. S. McDade D. Lallement and J. Saghir. (2005). Energy and the Millennium Development Goals. New York: Energy Sector Management Assistance Programme, United Nations Development Programme, UN Millenium Project, World Bank. Nfah, E.M. J.M. Ngundam and R. Tchinda. (2007). "Modelling of solar/diesel/battery hybrid power systems for far-north Cameroon", Renewable Energy, 32 (5):832-844. Painuly, J.P. H. Rao and J. Parikh. (1995). "A Rural Energy-Agriculture Interaction-Model Applied to Karnataka State", Energy, 20 (3):219-233. 18

Parikh, J. (1985). "Modeling Energy and Agriculture Interactions .1. A Rural Energy-Systems Model", Energy, 10 (7):793-804. Parikh, J.K. and R. Ramanathan. (1999). "Linkages among energy, agriculture and environment in rural India", Energy Economics, 21 (6):561-585. Pohekar, S.D. and M. Ramachandran. (2004). "Application of multi-criteria decision making to sustainable energy planning--A review", Renewable and Sustainable Energy Reviews, 8 (4):365-381. Ramirez Gomez, M. (2007). Pobreza y servicios publicos domiciliarios (in Spanish). . Bogota, Colombia: Deparatmento Nacional de Planeacion. Rubab, S. and T.C. Kandpal. (1996). "Biofuel mix for cooking in rural areas: implications for financial viability of improved cookstoves", Bioresource Technology, 56:169-178. Silva, D. and T. Nakata. (2008). "Renewable technologies for rural electrification in Colombia: a multiple objective approach", International Journal of Energy Sector Management, 2 (1):139-154. Silveira, S. (2005). Bioenergy - Realizing the potential. Oxford: Elsevier. Sims, R.E.H. (2002). The brilliance of bioenergy in business and in practice. London: James & James. Solantausta, Y. and J. Huotari. (1999). "Power production from wood - Comparison of the Rankine cycle to concepts using gasification and fast pyrolysis, IEA Bioenery task 22: technoeconomic assessments for bioenergy applications, Paris: IEA. Thankappan, S. P. Midmore and T. Jenkins. (2006). "Conserving energy in smallholder agriculture: A multi-objective programming case-study of northwest India", Ecological Economics, 56 (2):190-208. Underwood, C.P. J. Ramachandran R.D. Giddings and Z. Alwan. (2007). "Renewable-energy clusters for remote communities", Applied Energy, 84 (6):579-598. UPME. (2000). Establecimiento de un plan estructural, institucional y financiero, que permita el abastecimiento energético de las zonas no interconectadas con participación de las comunidades y el sector privado. Centros poblados, caracterización energética y agrupación. Bogota: UPME ———. (2003). Potencialidades de los cultivos energeticos y residuos agroindustriales en Colombia. Bogota: UPME ———. (2008). Boletin estadistico de minas y energia 2003-2008 (in Spanish). Bogota, Colombia. Bogota, Colombia: Unidad de Planeacion Minero Energetica. Urban, F. R.M.J. Benders and H.C. Moll. (2007). "Modelling energy systems for developing countries", Energy Policy, 35 (6):3473-3482. van Ruijven, B. F. Urban R.M.J. Benders H.C. Moll J.P. van der Sluijs B. de Vries and D.P. van Vuuren. (2008). "Modeling Energy and Development: An Evaluation of Models and Concepts", World Development, In Press, Corrected Proof. World Bank. (1996). Rural energy and development. Improving energy supplies for two billion people. Washington: World Bank. Yue, C.D. and S.S. Wang. (2006). "GIS-based evaluation of multifarious local renewable energy sources: a case study of the Chigu area of southwestern Taiwan", Energy Policy, 34 (6):730-742.

19