UEF//eRepository
DSpace https://erepo.uef.fi
Rinnakkaistallenteet Luonnontieteiden ja metsätieteiden tiedekunta
2018
A comparative study on household level energy consumption and related emissions from renewable (biomass)
and non-renewable energy sources in Bangladesh
Baul, T K
Elsevier BV
Tieteelliset aikakauslehtiartikkelit
© Elsevier Ltd
CC BY-NC-ND https://creativecommons.org/licenses/by-nc-nd/4.0/
http://dx.doi.org/10.1016/j.enpol.2017.12.037
https://erepo.uef.fi/handle/123456789/6652
Downloaded from University of Eastern Finland's eRepository
1
A comparative study on household level energy consumption and related emissions from renewable (biomass) and non-renewable energy sources in Bangladesh
T.K. Baula,b,⁎, D. Dattaa, A. Alamb
a University of Chittagong, Institute of Forestry and Environmental Sciences, Chittagong 4331, Bangladesh
b University of Eastern Finland, Faculty of Science and Forestry, School of Forest Sciences P.O.
Box 111, FI-80101 Joensuu, Finland Abstract
In developing countries, securing clean and equal energy access for all is often constrained by lack of understanding of households’ energy dependency and influencing factors. This study investigates household-level energy consumption patterns, relevant socioeconomic factors and carbon-emissions from various energy sources. Using a semi-structured questionnaire, we conducted an explorative survey of 189 households in three income groups in a suburban region of Chittagong, Bangladesh.
Results suggest that most of the households heavily depend on biomass energy that accounts for 87% of their monthly energy consumption and about two-thirds of energy expenditure.
Contrariwise, dependence on non-renewable energy is minimal and accounts for households’ 31%
monthly energy expenditure. The rich households tend to rely more on electricity, candle, liquid petroleum gas (LPG) while their consumption of the non-renewables is significantly higher than that of medium-income and poor households. Income, education and landholdings of households are positively correlated with expenditure for consuming convenient energy sources such as firewood, electricity and LPG. Firewood, the biomass fuel used most for cooking, is an energy source with the highest carbon emissions—monthly about 192 kilogram carbon dioxide equivalent per household.Our research findings offer insights to enhance household-level clean energy access in Bangladesh and countries alike.
Keywords: biomass; emission; energy expenditure; firewood; income; non-renewable
2 1. Introduction
Non-renewable and renewable energy sources constitute a percentage of about 81 and 14, respectively, of the global total primary energy supply (IEA, 2017). The share of non-renewable fuels include oil, coal, and natural gas, while renewable energy sources (RES) include biomass, sunlight, wind, tides, waves, hydro and geothermal. In 2015, the percentage of renewable energy in global final energy consumption was about 14% (IEA, 2017), of which traditional biomass contributed with about 9% as one of the major RES, especially in the developing countries (REN21, 2016). Globally, the proportion of biomass energy will reach 50% by 2050 in terms of consumption (Mondal and Denich, 2010). Biomass, a combination of different organic compounds, is mainly derived from three sources: agricultural residues, forest residues, and energy crops (Guta, 2012).
Generally, biomass refers to rice husk, crop residues, jute sticks, wood, leaves and forest residues, animal waste, municipal waste, etc (Hossen et al., 2017). Conversion of biomass into bioenergy for the production of heat and electricity occurs via two widespread technologies: direct combustion and gasification (Mondal and Denich, 2010), which play vital roles in the substitution of non- renewable fossil fuels. Locally available traditional forms of biomass are used via direct combustion mostly in rural areas of the developing countries. However, increased use of biomass in an efficient way via improved technology can potentially contribute to a clean environment by reducing emissions and representing a promising source of electricity and gas (Hossen et al., 2017).
1.1 Bangladesh energy sector and potential of biomass
In Bangladesh, the demand for energy (with an annual growth rate of 10%) are currently not being met (GOB PD, 2011). According to the World Bank (2014), the per capita energy use in Bangladesh was 215.52 kilograms of oil equivalent (kgoe) in 2013, which was very low compared to those of its neighboring countries, such as, India, Pakistan, Srilanka, and Bhutan. In Bangladesh, only 61% of the population have access to electricity, with a per capita consumption of 293.03 kWh
3
a-1 (kilowatt hour per year) (REN21, 2016; World Bank, 2014). About 91% of the country’s total electricity generation depend on non-renewable energy sources (natural gas, furnace oil, diesel, coal), with natural gas being the largest contributor (69%) (BPDB, 2015). Nevertheless, the demand for electricity is higher than the production throughout the country (Islam et al., 2014a; Mondal and Denich, 2010). The government of Bangladesh (GOB) commenced importing crude oil and electricity from other countries, albeit at high costs (Amin et al., 2016). Bangladesh has spent almost 5 billion US$ to import 5,450 thousand tons of crude oil and petroleum products in 2014;
additionally, approximately 7% of the overall electricity supply were imported in the same year (BBS, 2014; BPC, 2015; BPDB, 2015). The GOB aims to provide electricity for citizens by 2020 (GOB PD, 2011; Power System Master Plan, 2016). Natural gas accounts for 75% of the primary energy consumption and has long been used in industries, fertilizer factories, and in domestic and transport sectors (BP, 2013). However, the reserves of natural gas and coal are limited in comparison to the development needs of the country (Ahmed et al., 2013; Huda et al., 2014). Thus, within the next few decades, Bangladesh will be aiming to confront the serious energy crisis caused by over-dependence on non-renewable fossil fuels (Ahmed et al., 2014; Islam et al., 2014a).
Renewable energy sources represented only about 1% of the total electricity generation in 2015 (BPDB, 2015), which the GOB has envisioned to reach up to 10% by 2020, as declared in the Power System Master Plan 2016 of Bangladesh (GOB PD, 2011; Power System Master Plan, 2016;
REN21, 2016). Among the renewable energy sources, solar and biomass energy may have the highest potentials (e.g. Ahmed et al., 2013; Amin et al., 2016; Islam et al., 2006; Mondal and Denich, 2010). Traditional biomass fuels are predominant sources of rural energy, contributing over 90% to the total primary energy supply (BBS, 2010; Mainali et al., 2014) to meet cooking, commercial, and industrial needs (Rahman et al., 2013), mainly in the form of agricultural residues (46%), wood wastes (34%), and animal dung (20%) (Huda et al., 2014; Islam et al., 2014b).
However, biomass as a source of clean energy to avoid traditional uses appears worth mentioning
4
when integrated with the Clean Development Mechanism (CDM) project, which entails using improved cooking stoves, thus reducing carbon dioxide (CO2) emissions and deforestation, as well as generating employment and income supports (Miah et al., 2009; Uddin and Taplin, 2009).
Over 70% of the population live in rural areas (World Bank, 2013), where income limitations prevent rural households from accessing convenient forms of energy. All rural areas are not well developed and lack access to modern facilities and the electrification network (Islam et al., 2014a; Mondal and Denich, 2010); therefore, the rural population depends on kerosene and candles for lighting. By contrast, rural areas with electrification, but without access to the national gas supply network, rely on the use of biomass for cooking and heating (Ahiduzzaman and Islam, 2011;
Halder et al., 2014). Additionally, rural poor people often cannot afford buying liquefied petroleum gas (LPG) for cooking. However, rural people’s consumption pattern of traditional biomass fuel is regulated by socioeconomic factors such as literacy of household and availability and cost of biomass, and resources around them (Jumbe and Angelsen, 2011; Rao and Reddy, 2007). For example, home garden (or homestead forests) occupy 0.3 million ha of land (12% of the total forest cover in the country) and play a potential role in providing wood fuels and forest residues (BFD, 2016). Conversely, the rural people of northern regions in Bangladesh, where forests are sparse, mainly rely on agricultural residues, including rice straw, rice husk, rice bran, jute stalk, dung cake, etc. (Halder et al., 2014; Hassan et al., 2014; Huda et al., 2014; Jashimuddin et al., 2006).
1.2 Objective (s) of the study
Several studies have investigated the energy-use pattern in different rural areas of Bangladesh and found variation in energy consumption (e.g. Foysal et al., 2012; Hassan et al., 2012; Miah et al., 2010, 2011a). At the household level, the study by Akhter et al. (2010a), Jashimuddin et al. (2006) and Miah et al. (2003) found that traditional use of biomass energy had a significant contribution to the rural energy supply in the southern and central part of Bangladesh.
Others documented the shortage of biomass, specifically firewood in forest-rich and degraded areas,
5
due to overexploitation of the resources (Akther et al., 2010b, 2010c; Hassan et al., 2014; Miah et al., 2009; Nath et al., 2013). However, according to Hassan et al. (2012), little information is available on rural households’ energy consumption patterns and their expenditure on various types of energy, quantity, and sources of fuels. For instance, Miah et al. (2003) did not consider natural gas and other non-biomass fuels in the study conducted in Chittagong region. Moreover, studies on village level energy consumption patterns in regions close to urban areas are scarce. Energy consumption patterns vary with the locations of the study area: urban, suburban, and rural and the proximity to forest areas (Heltberg, 2005; Miah et al., 2011a; Rahut et al., 2014). The present study represents villages with heterogeneity in energy supply. Furthermore, our study area represents the heterogeneous features of the households in terms of income and landholdings, which is a vital aspect to study socioeconomic factors as driving forces of energy consumption and expenditure (Akther et al., 2010a; Behera et al., 2015; Heltberg, 2005; Ouedraogo, 2006; Pachauri, 2004; Reddy and Srinivas, 2009). The emission reductions in using traditional biomass fuels were emphasized;
therefore, study taking into account the CO2 emissions from using biomass energy need to be carried out for the potential development of bioenergy in Bangladesh (Miah et al., 2011b).
From the above-discussed studies, it is evident that there is a dearth of comprehensive information regarding consumption and expenditure of renewable (biomass) and non-renewable energy and their comparisons in rural Bangladesh. Furthermore, no research was so far conducted on CO2 emissions released from the usage of energy fuels in rural and suburban areas of Bangladesh. In addition, how socioeconomic factors of the households influence the expenditure for energy consumption at household level were not properly reflected in previous studies. Therefore, our key research question addresses “What is the pattern and variation in household consumption and expenditure of non-renewable and biomass energy among different income groups?” Based on this question, we aimed at investigating household level consumption and expenditure of non- renewable and biomass energy, along with their sources and end uses, and related socioeconomic
6
factors (household size, income, literacy, and landholdings) under three income groups in a suburban region of Chittagong, Bangladesh. We also studied the CO2 emitted from the monthly consumption of the biomass and non-renewable energy fuels. Our goal was to provide policy makers, researchers, and concerned stakeholders with an energy resource base, especially in the context of sustainable energy supply for the development of rural Bangladesh and countries alike.
2. Materials and methods
2.1 Study site
The study was carried out at Hathazari Upazila1 of the Chittagong District, located between 22°24´
and 22°38´ N latitude and 91°41´ and 91°54´ E longitude, with an area of 25,500 ha (Fig. 1). The region includes 3,252 ha of public forest and 17,665 ha of cultivated land (BBS, 2013). In the South, the Upazila borders to Bayjid Bostami and Chandgaon Thanas, which are under the jurisdiction of the Chittagong City Corporation. Geographically, the area consists of small hills with poor stocks of public forests and private plantations (woodlot).
Total population of the Upazila is about 431,748, with approximately 81,292 households in which males and females are equally distributed, with an average literacy rate of 65% (BBS, 2013).
In the Upazila, approximately 46% of the total population is landless (who only have dwelling land). The population mainly depends on agriculture, local businesses (small tea or grocery shops), and remittances (BBS, 2013). All the Unions of the Upazila are under the rural electrification network, but only 69% of the households, mainly those near the Union, have access to the electricity grid. This Upazila has an uneven distribution of electricity and natural gas (Upazila
1In the governance system of Bangladesh, Upazilas (sub-districts) and Unions are regarded as the most important local government strata and social institutions. An Upazila consists of a few Unions, which are composed of many villages and/or wards. Village and ward are the two lowest units of the local government system, and their activities are governed by the Union office.
7 Parishad Office, 2014).
Fig. 1. Maps of Bangladesh (right panel) and Hathazari Upazila (left panel); study areas are highlighted by circles. Source: Banglapedia
2.2 Sampling and data collection 2.2.1 Sampling strategy
The sequence of the sample selection procedure was from Upazila to Union, from Union to village, and from village to household level. In the Hathazari Upazila, three out of sixteen Unions, namely Mirzapur, Fatehpur, and Hathazari were selected based on the availability of public forests within the Union (Table 1). In the selection of Unions, taking the availability of public forests into account was likely to represent the area with heterogeneity in energy supply, although villagers usually have no access in these public forests to collect biomass. The forest area is under the administrative
8
jurisdiction of the Hathazari Forest Beat2 of the Bangladesh Forest Department (Hathazari Forest Beat, 2014). A reconnaissance survey was conducted in those three Unions to obtain an overview of the whole study area. Following Hassan et al. (2013), the number, location, and socioeconomic status of the villages were collected from these Union council offices and Upazila Parishad office.
Three villages were selected from each of the three Unions based on their location (e.g.
nearby: 1-2 km, far: 3-4 km, and very far: > 4 km from the public forest) (Table 1). Thus, the selected nine villages represented the communities of the Upazila (Table 1). The numbers and lists of the households and some brief ideas of their income status for each of the selected villages were obtained from the respective Union council and Upazila Parishad office. Then, a meeting was held with key informants such as elderly and knowledgeable persons in the presence of the heads of the villages to cross check the data about households and their income status obtained from the respective Union council offices. In this process, four key informants were selected from four different parts of a village, and we assumed that village heads and informants know all households of their respective villages. Based on their consensus in the meeting, all households were categorized into three income groups: rich (> 195 US$/month), medium (91–195 US$/month), and poor (< 91 US$/month), which was further verified by the pre-tested questionnaire. Income of the household referred to as the earnings from on-farm (e.g. agriculture, woodlot, homestead) and off- farm (e.g. services, business, day labor) activities. Finally, 21 households from each village (7 households from each income group), comprising a total of 189 households (n = 189; 7 households
× 3 income groups × 3 villages × 3 unions), were selected by using a random number table (Table 1). Subsequently, the names and locations of the households were checked and identified with the help of a respective local guide from the Ward and Union council office
2Beat is the ultimate and lowest administrative field unit of forest management of the Bangladesh Forest Department (BFD).
9
Table 1. Selected Unions and villages along with households for surveying
Unions Area (ha) a
Forest area (ha) a
Villages Distance from public forest b
No of household b
Households surveyed Mirzapur 2,611 1182 Hasimnagar 2 km south west
(nearby)
299 21 (R=7; M=7; P=7) c Charia 4 km west
(far)
3184 21 (R=7; M=7; P=7) Mirzapur 7 km south west
(very far)
3587 21 (R=7; M=7; P=7) Fatehpur 2,413 134 Jungle
poschimpotti
1 km west (nearby)
428 21 (R=7; M=7; P=7) Mithachora 3 km west
(far)
585 21 (R=7; M=7; P=7) Maizpotti 5 km west north
(very far)
2108 21 (R=7; M=7; P=7) Hathazari 2,294 508 Jungle
shilchhari
1 km west (nearby)
48 21 (R=7; M=7; P=7) Chandrapur 3 km west
(far)
330 21 (R=7; M=7; P=7) Mirarkhil 6 km west
(very far)
928 21 (R=7; M=7; P=7)
a BBS (2013)
b nearby: 1-2 km, far: 3-4 km, and very far: > 4 km from the public forest. Source: Union Council Office (2014)
c R: Rich; M: Medium; P: Poor income group< Table 1 >
2.2.2 Pre-testing of the questionnaire
A pre-testing of the questionnaire was conducted with 10 households to ensure its clarity3, comprehensiveness4, and acceptability5 of the questions for the respondents (Rea and Parker, 1997).
In addition, the key informants such as village head, elderly, and knowledgeable persons were inquired about the acceptance of the questionnaire. A snowballing6 approach was employed to identify the appropriate informants and households for pre-testing and further information. Based on the feedback received from the respondents and key informants, the questionnaire was modified and reformulated. This pre-testing also enabled us to adjust the data collected from the Union
3Are the questions understood by the respondents?
4Are the questions and response choices sufficiently comprehensive to cover a reasonably complete range of alternatives?
5Is there any problem with questionnaire length or questions that are perceived to invade the privacy, moral and ethical standards of respondents?
6 a simple technique asking a key informant to name other target people who are important in the study in relevant aspect such as income, land holdings, expenditure for households (Narayan, 1996).
10
council office about the socioeconomic status of the villagers.
2.2.3 Data collection
The researchers, in person, interviewed the head of each household, regardless of gender, with the presence of local guide by using a semi-structured questionnaire consisting of both open- and close- ended questions. In absence of the households’ heads, the next to the heads of the households were interviewed. In the case of female respondents, the local female guide facilitated the interview. The interview was mainly based on memory recall and estimates. The questionnaire consisted of 20 questions, which were divided into two parts to collect information on socioeconomic and energy usage of the respondents. Socioeconomic data included landholdings, income, education, and size of the households. Energy usage data included the types, sources, and end uses of energy consumed, including total monthly consumption and expenditure on both renewable and non-renewable fuels.
Regarding the overestimation of biomass consumption, data provided by the respondents were randomly verified through spot measurement and direct observation. With respect to expenditure on fuels, the respondents were requested to provide their actual monthly expenses (e.g., from the monthly invoice of electricity and natural gas) for purchasing renewable and non-renewable fuels.
The respondents were also asked to express their opinions on what was not asked. This way, new avenues of questioning were developed while interviewing. Field work was conducted in several weeks over a period of April to July 2014.
2.3 Data analysis
The collected data on the socioeconomic status were analyzed and expressed in mean values for different landholdings, income, literacy, and size of the households. Moreover, respondents’
occupations were analyzed and expressed in percentage (%) of the respondents. We also calculated and discussed the contributions (in percentage, %) of occupations and literacy on the income of the
11
households under the various income groups. With regard to energy, the percentage (%) of the households was analyzed for items such as renewable and non-renewable energy users with fuel types and their sources and end uses. Subsequently, energy consumption and expenditure of the households were analyzed and expressed in mean monthly values. We also calculated the expenditure ratio (in %) for energy type (ET) to the total energy (TE) and to the total income (TI) for both renewable and non-renewables under the various income groups. The comparison among three different income groups of the households was performed in terms of their landholdings, income, end uses, sources, monthly consumption, and expenditure of energy fuels. One-way analysis of variance (ANOVA) was used to test the significance for mean values of these variables and to evaluate the significant variation between income groups. Duncan’s Multiple Range test (DMRT) was applied to measure which groups are significantly different from which. In addition, Spearman's correlation was applied to determine the relationships between energy expenditure and income, landholdings, literacy, and size of the households. For all statistical analyses, we used the package SPSS 17.0. and R. The details on how the primary data were made comparable are given below.
2.3.1 Literacy
The educational values were coded based on the usual time span of the degree awarded in Bangladesh (adopted from Miah et al., 2010). The used values were 0 for illiterate, 5 for primary education, 10 for secondary education, 12 for higher secondary education, 16 for graduate, and 19 for post-graduate degree holders. To calculate the weighted score of literacy per household, the education values of all members in a household were summed up and then divided it by the total family members, excluding children five years and younger.
2.3.2 Non-renewable and renewable energy
In this study, the non-renewable energy fuels included kerosene, liquefied petroleum gas (LPG),
12
candles, natural gas, and grid electricity. Currently, about 91% of the grid electricity is obtained from non-renewable sources, mainly from natural gas, furnace oil, diesel, and coal, and 1% is derived from renewables, mainly hydropower; the rest is imported (BPDB, 2015). During the interview, the units used for non-renewables were kilogram (kg) for candles, liter (l) for kerosene and LPG, cubic feet (ft3) for natural gas, and kilowatt hour (kWh) for electricity. Thereafter, all these units were converted into a single uniform unit (kWh) by using the energy content of the corresponding fuels available from the literature (Table 2). For example, monthly consumption of kerosene per household recorded in liter (l) during the interview was converted into kWh by multiplying with its heating value per unit (9.8 kWh l-1).
Table 2. Heating value and emission factors used for non-renewable fossil and biomass fuels.
Non- renewables
Heating value a Emission factor a (kg CO2 e kWh-1)
Biomass Heating value b (kWh kg-1)
Emission factor c (kg CO2 ekg-1)
Electricity - 0.619 Firewood 4.2 1.163
Kerosene 9.8 kWh l-1 0.298 Bamboo 5.35 -
Liquefied petroleum gas
6.66 kWh l-1 0.241 Dry leaves and twigs
3.5 -
Candles 11.67 kWh kg-1 0.25 Crop residues 3.5 1.174
Natural gas 0.29 kWh ft-3 0.185 (cooking) Dung cake 2.4 0.787
Source: a IEA (2010a), IEA (2010b) b Biswas and Lucas (1997), Islam (1980)
c Bhattacharya et al. (2000)
The major renewable energy in the study area was based on biomass. Thus, the renewable energy implied different types of biomass fuels, which included firewood, leaves and twigs, bamboo, crop residues (rice husk, rice straw, and other crop residues), and dung cake. Data on the biomass were collected using the local measuring units, i.e. auri, maund, cartload, or headload; such units have specific relations to kilogram (kg) by weight. Auri can be range between 25 and 35 kg, maund means 37 kg, cartload ranges from 700 to1000 kg and headload from 20 to 25 kg. Variation in the mass, for example in cartload and headload, also dependent on the types of biomass and loading capacity of the transportation means. Since there is variation in the units, one standard unit
13
was used at every Union, calculated by verifying the local measuring units and converted into kilogram (kg). Afterwards, the mass of the biomass (kg) was converted to kilowatt hour (kWh) by multiplying with the corresponding energy content (Table 2). The moisture contents of firewood and dung cake were assumed to be 15-30% and 50%, respectively (Hossain, 2003; Miah et al., 2009). The remaining biomass fuels were assumed to be air-dried matter because they were physically checked and verified by the user groups before burning.
2.3.3 Emission calculation
Carbon dioxide (CO2) emissions from the monthly consumption of the biomass and non-renewable energy fuels were derived from Equation (1), adopted and modified from Bhattacharya et al. (2000).
The results of the emissions were expressed in kg CO2 e household-1 month-1.
(1),
where, EF is the emission factor of the fuel (in kg CO2 e kg-1 or kg CO2 e kWh-1 of the fuels) (Table 2), activity is the amount of fuel consumed (kg or kWh), a and b denote the types of energy fuel and activity, respectively.
2.3.4 Currency conversion
The primary data on energy expenditure was recorded in Bangladesh currency (BDT) and converted into US dollars (US$) by using the conversion rate of US$ 1= BDT 77 (date of relevance: July 2014).
2.3.5 Energy-energy and energy-income ratio calculation
The expenditure ratios for energy type (ET) to total energy (TE) and energy type to total income (TI) of the households were calculated by applying Equation (2). For example, monthly expenditure
14
ratio for firewood was calculated with respect to total energy expenses and total income of the household. The ratio was then converted into percentage value and reported in the Results and discussion section.
(2),
where ET, TE, and TI indicate the types of energy, total energy used, and total income generated by the household in one month.
3. Results and discussion
3.1 Socioeconomic profile of the households
Across the income groups, business was found to be the main occupation, representing 44% of the total respondents. Others were engaged in farming (18%), service (16%), day labor (15%), fishing (6%), and teaching (1%) (Table 3). In addition to the main occupation, 13% of the respondents were engaged in secondary occupations, such as day labor, business, fishing, and farming.
Table 3. Percentage (%) of the main occupations of the respondents and literacy of the households’
members by the income groups.
Main occupation of the respondents
Rich Medium Poor Total Literacy of the household members
Rich Medium Poor Total
Business 20 14 11 44 Higher secondary 5 0 0 5
Farming 6 6 6 18 Secondary 24 18 3 45
Teaching 1 0 0 1 Primary 4 15 29 48
Fishing 0 2 4 6 Illiterate 0 0 2 2
Day labor 0 5 10 15
Service 6 7 3 16
The size of the household ranged between 3 and 11 members, with 6 members on average,
15
regardless of the income groups (rich, medium, and poor). Males and females were equally distributed among the households. The mean educational score of the family members was 10.7, which indicates an average education of up to secondary level. Rich households had education at the level of higher secondary (score 12.10) and secondary, while poor households had only a primary education level (Table 3).
Considering all sources, such as on-farm (e.g. sale of agricultural crops and tree products) and off-farm (e.g. business, service) activities, the mean monthly household income was 169.93 US$, with a maximum of 1,558.44 US$ and a minimum of 51.95 US$. Mean monthly income values for rich, medium-income, and poor households were 308.39 ± 26.14, 129.05 ± 2.59, and 72.36 ± 1.30 US$, respectively. This was due to the fact that business contributed the major share of income (60%) of the rich households, while it was 41% for medium-income households (data not shown).
The poor households’ major income sourced from business such as small shop, tea stall, and day labor and these were 35% and 29%, respectively (data not shown). The households with the large landholdings also had a larger income. As expected, the rich households had the largest share of landholdings, regardless of the type of land (Fig. 2), and the rich and medium-income had 302%
and 73% larger landholdings, respectively, compared to that in the poor households. Regarding the literacy, the rich households with higher secondary and secondary education contributed 22% and 66%, respectively, of their incomes. The major share of the medium-income (57%) and poor (85%) households’ income came from those who had secondary and primary education, respectively (data not shown).
16
Fig. 2. Mean areas of homestead, agricultural land, and woodlot under various income groups. Bars represent standard error of mean.
3.2 Users, consumption and end-uses of non-renewable energy
In our study, most households used more than one variety of non-renewable fuel for their everyday domestic energy needs. Grid electricity (73% of households) and kerosene (72%) were the first and second most common energy sources used for lighting, followed by candles (27%) for lighting, by LPG (13%), and natural gas (5%) for cooking. Nearly all rich households (98%) used electricity, while 86% and 83% of the medium-income and poor households, respectively, used kerosene.
The mean monthly consumption of total non-renewables was 166.03 kWh household-1 in the study area. Amongst the non-renewables, natural gas dominated, with 77.33 kWh household-1, while candles only accounted for 0.63 kWh household-1. Regarding the income groups, the rich households’ monthly consumption of total non-renewable energy was significantly higher than those of the other two income groups, while no significant difference was found between poor and medium income groups. In particular, electricity consumption between rich (96.73 kWh), medium- income (45.35 kWh), and poor (16.19 kWh) households significantly (p < 0.05) differed. In the case
17
of LPG and candles, the rich households’ consumption was significantly (p < 0.05) higher than that of the other two groups. On the other hand, the medium-income households’ kerosene consumption was highest, followed by those of the poor and rich households, respectively, with significant (p <
0.05) differences (Table 4).
Table 4. Mean monthly consumption of non-renewable energy (kWh household-1) with standard error of mean (±) by the income groups. Superscripts within a column (a, b or c) indicate significant differences at p < 0.05 based on the income groups. N represents the number of the respondents.
Income group Electricity Kerosene LPG Candles Natural gas Mean of total Rich (N = 63) 96.73a ± 7.14 8.17a ± 1.11 50.43a ± 12.11 1.37a ± 0.25 232.00 ± 67.83 388.69a ± 65.99 Medium (N= 63) 45.35b ± 3.71 17.27b ± 1.44 17.02b ± 6.11 0.35b ± 0.11 0.00 ± 0.00 79.99b ± 6.69 Poor (N = 63) 16.19c ± 2.57 13.07c ± 1.01 0.00 ± 0.00 0.17b ± 0.06 0.00 ± 0.00 29.43b ± 2.25 Mean (N = 189) 52.76 ± 3.71 12.83 ± 0.74 22.48 ± 4.75 0.63 ± 0.10 77.33 ± 23.86 166.03 ± 24.86
In our study area, the electricity distribution was uneven due to insufficient rural electrification, higher connection and access costs, and bureaucratic complexity, which were the major reasons for the prominent use of kerosene. This shows that the rich could afford to pay the high costs of a grid-electricity connection. The households in all income groups with access to grid- electricity used candles and kerosene lamps for lighting purpose when power disruption occurred;
however, the rich households preferred candles instead of kerosene. Thus, the dependency on kerosene and candles had increased due to increasing power disruption. The overall monthly consumption of electricity, kerosene, and candles matched those observed in an earlier study by Miah et al. (2010). However, electricity consumption was higher than that (12 kWh) reported in Asaduzzaman and Latif (2005) and 29.63 kWh in Hassan et al. (2012), which might be due to uneven electrification in very remote areas. Natural gas and LPG were used for cooking by rich and medium-income households, who sometimes used kerosene to facilitate the catching fire for cooking, especially in the rainy season. None of the poor households used natural gas and LPG, which is most likely due to inaccessibility and expensiveness, respectively. Such a type of energy consumption pattern is consistent with previous findings in the context of rural Bangladesh (e.g.
Asaduzzaman et al., 2010; BBS, 2010; Hassan et al., 2012; Miah et al., 2010).
18
3.3 Users, consumption, sources and end-uses of biomass
On average, 92% of the households across the three income groups used biomass for cooking, followed by that for rice paraboiling, water heating, and as compost and animal feed. Biomass was the leading fuel for cooking and used by all poor and the majority of the medium-income and rich households (Table 5). The households used more than one variety of biomass fuels, such as firewood, leaves and twigs, bamboo, crop residues, and dung cake recorded in this study. Firewood was the dominant biomass type, while crop residues (rice husk, rice straw, and other crop residues) were the least in terms of user numbers. Over 80% of the households in both rich and medium income groups used firewood, while 85% of the poor households used leaves and twigs. Only poor (30%) and medium-income (17%) households used dung cake. Bamboo was used by 19-35% of the households, irrespective of the income groups (data not shown).
Table 5. End uses of biomass by household income groups. Values in parenthesis indicate actual number and outside parenthesis indicate percentage of the households. N represents the number of the respondents. a Others include livestock feeds, compost.
Income group Cooking Rice parboiling Water heating Others a Rich (N = 63) 82.54 (52) 6.34 (4) 6.34 (8) 7.94 (5) Medium (N = 63) 93.65 (59) 14.28 (9) 4.76 (3) 3.17 (2) Poor (N = 63) 100.00 (63) 3.17 (2) 0.00 (0) 0.00 (0) Mean (N = 189) 92.06 (174) 7.93 (15) 5.82 (11) 3.70 (7)
Mean monthly consumption of total biomass was 1087.79 kWh household-1 (equivalent to a mass of 165.08 kg) in the study area. Amongst the biomass fuels, the consumption of firewood was highest (693.27 kWh household-1), followed by that of leaves and twigs, bamboo, dung cake, and crop residues, respectively. Regarding the income groups, no significant differences (p < 0.05) were found in the consumption of total biomass by rich (1,160.97 kWh or 288.25 kg), medium-income (1,142.58 kWh or 250.19 kg), and poor (959 kWh or 217.78 kg) households. However, significant
19
differences were found for some types of biomass fuels mainly between the poor and the other two income groups. Firewood consumptions of the rich and medium income groups were significantly higher (p < 0.05) than that of the poor. However, the poor households’ consumptions of leaves and twigs, as well as dung cake, were significantly higher (p < 0.05) compared to the other two groups (Table 6).
Table 6. Mean monthly consumption of various biomass fuels (kWh household-1) with standard error of mean (±) by the income groups. Different superscripts (a, b or c) within a column indicate significant differences at p < 0.05 based on the income groups. N represents the number of the respondents.
Income group Firewood Leaves and twigs
Bamboo Crop residues
Dung cake Mean of total Rich (N = 63) 887.92a ± 69.41 192.71a ± 22.20 78.12a ± 17.66 2.22a ± 2.22 00.00 ± 0.00 1160.97a ± 97.39 Medium (N = 63) 834.59a ± 61.76 241.45a ± 24.45 44.16a ± 12.96 0.00 ± 0.00 22.39a ± 6.49 1142.58a ± 55.23 Poor (N = 63) 357.30b ± 43.09 462.96b ± 20.90 68.36a ± 15.91 0.02a ± 0.01 66.78b ± 10.95 959.82a ± 35.14 Mean (N = 189) 693.27 ± 38.15 299.04 ± 15.54 63.54 ± 9.04 2.22 ± 1.27 29.72 ± 4.68 1087.79 ± 39.47
In this study, firewood use increased with increasing household income, replacing the use of leaves and twigs, and dung cake. Thus, the consumption of firewood and leaves and twigs by the rich, medium-income, and poor households constituted the lion´s share of the total biomass consumption, which is consistent with findings obtained in rural areas of Bangladesh (Asaduzzaman and Latif, 2005). Similar patterns of firewood use have also been reported in earlier studies (e.g. Balat, 2009; Hassan et al., 2012, 2013; Jashimuddin et al., 2006; Miah et al., 2010).
However, in the present, the monthly dominant consumption of firewood (693.27 kWh household-1) differed to that found by Akther et al. (2010a), who stated that consumption of leaves and twigs dominated (672.20 kWh household-1; converted from their original mass), followed by firewood, in the central region of Bangladesh. This discrepancy most likely occurred because of the difference in the variety of biomass in different regions of the country, which thus resulted in a variation in the consumption pattern. However, a higher collection of leaves and twigs, and dung cake by the poor
20
households may decrease the fertility of forests and agricultural land. Contrarily, only a few medium-income (3%) and rich (8%) households used leaves as well as crop residues as livestock feed and crop residues and dung cake as compost to maintain land fertility (Table 5). Thus, the selection of biomass fuels was regulated by the household pattern. The FAO (2009), in line with our study, revealed that biomass is the cheapest and most accessible source of cooking fuel for the majority of rural people in the Asian and Pacific region.
Regardless of the income groups, biomass supply was mainly, for around 70% of the households, dependent on the homestead forest, as illustrated in Fig. 3. This is in agreement with the findings from Hassan et al. (2012) and Nath et al. (2013). However, extensive dependency on the homestead forest for biomass collection may lead to the gradual erosion of the resources with a shortage of fuels in near future. In Bangladesh, homestead forests occupying 0.3 million ha of land (BFD, 2016), with a potential source of biodiversity and diverse foods (Baul et al., 2015; Nath et al., 2014; Roy et al., 2013), need to be saved considering its future demand and declining resources.
This requires sustainable forest management and practicing less harvest to save from depletion.
Finding out alternative sources of biomass fuels, such as woodlots (forest plantation) and markets can minimize over-dependency on homestead forest. In this study, about 43% of the rich and 16%
of the medium-income households purchased firewood from the market, whereas the poor households (37%) could buy firewood and other forms of biomass cheaply from the neighborhood (Fig. 3), in line with Miah et al. (2010). Thus, the lower purchasing power of the poor households prevented them from buying firewood at the market. The dependency of merely 24% of the poor households on biomass supply was from public forests. Within this, about 87% households collected biomass fuels from nearby forests and the remaining 13% collected from far forests (data not shown). However, our findings differed from that of Islam and Sato (2012), where 72% of the poor households collected firewood from the Sal forest in Bangladesh in an uncontrolled and illegal manner. Conversely, the rich and medium-income households, due to their purchasing ability, were
21
not dependent on nearby forests for biomass collection, thereby preventing further degradation of public forests. Dissimilar to this study, the people from regions with insufficient forests were dependent on agricultural land for crop residues in India (Hiloidhari et al., 2014).
Fig. 3. Sources of biomass against the percentage of the households under various income groups.
3.4 Household energy expenditure
Mean monthly expenditure for total energy (non-renewables and biomass) consumption was 23.96
± 0.73 US$ household-1, of which biomass contributed more than twice as much compared to the non-renewables. Regarding the biomass fuels, the households spent the highest amounts on firewood (10.05 ± 0.57 US$) and lowest on dung cake (0.95 ± 0.15 US$). With respect to non- renewables, the expenditure was highest for electricity (3.10 ± 0.22 US$) and lowest for candles (0.14 ± 0.02 US$). In terms of income groups, households with higher incomes spent significantly more money on energy (p < 0.05) than the households with lower incomes, especially for electricity, kerosene, LPG, candles, and firewood. In contrast, the expenses for leaves and twigs as
22
well as dung cake were significantly higher in the lower income group than in the relatively higher income group. The rich, medium-income, and poor households’ expenditures for total energy consumptions accounted for 31.78 ± 1.62, 22.37 ± 0.72, and 17.71 ± 0.28 US$, respectively (Table 7).
Table 7. Mean monthly expenditure (US$ household-1) for biomass and non-renewable energy with standard error of mean (±) by the income groups. Different superscripts (a, b or c) within a column indicate significant differences at p < 0.05 based on the income groups. N represents the number of the respondents.
Non-renewables
Income group Electricity Kerosene LPG Candles Natural gas Mean of total Rich (N = 63) 5.68a ± 0.42 0.82a ± 0.11 6.06a ± 1.45 0.30a ± 0.05 0.90 ± 0.26 13.75a ± 1.41 Medium (N = 63) 2.66b ± 0.22 1.68b ± 0.14 2.02b ± 0.73 0.09b ± 0.03 0.00 ± 0.00 6.45b ± 0.72 Poor (N = 63) 0.95c ± 0.15 1.30b ± 0.10 0.00 ± 0.00 0.04b ± 0.01 0.00 ± 0.00 2.29c ± 0.13 Mean (N = 189) 3.10 ± 0.22 1.27 ± 0.07 2.69 ± 0.57 0.14 ± 0.02 0.30 ± 0.09 7.5 ± 0.63 Biomass
Income group Firewood Leaves and twigs
Bamboo Crop residues
Dung cake Mean of total Rich (N = 63) 14.22a ± 1.16 2.92a ± 0.34 0.89a ± 0.18 0.00 ± 0.00 0.00 ± 0.00 18.03a ± 1.55 Medium (N = 63) 11.13b ± 0.76 3.50a ± 0.42 0.61a ± 0.17 0.00 ± 0.00 0.69a ± 0.21 15.92 a ± 0.62 Poor (N = 63) 4.78c ± 0.55 7.27b ± 0.49 1.19a ± 0.24 0.00 ± 0.00 2.18b ± 0.36 15.42a ± 0.31 Mean (N = 189) 10.05 ± 0.57 4.56 ± 0.28 0.90 ± 0.12 0.00 ± 0.00 0.95 ± 0.15 16.46 ± 0.57
Table 7 also shows that there were no significant differences in biomass expenditures between income groups with monthly expenses over 15 US$ each. This was due to the availability of easily accessible biomass around the households. Consequently, the expenditure per household for biomass was 69% of the total energy expenditure (Table 8), which was lower, 38% and 36% as reported in Asaduzzaman et al. (2010) and Hassan et al. (2012), respectively. However, the expenditure ratio for biomass to the households’ total energy expenditure increased from the rich (57%) to the poor (87%) households (Table 8). Due to higher prices of firewood, the poor households had to gather and buy lower-grade biomass fuels, including leaves and twigs, for which they spent most of their energy budget (Biswas et al., 2011). In the case of non-renewables, we found significant differences in expenditure between income groups (Table 7). The prices of non-
23
renewables, especially electricity, LPG, and kerosene, have considerably increased over the years, making those energy sources unaffordable for the lower income group. Consequently, the expenditure ratio for the non-renewables to the households’ total energy expenditure decreased from the rich (43%) to the poor (13%) households (Table 8). The rich households spent more money of their total energy expenditures on good quality fuels, for example firewood, electricity, and LPG. Thus, the households’ monthly expenditure for total energy consumption was higher than those found in different agro-ecological zones of Bangladesh (Foysal et al., 2012; Hassan et al., 2012).
Table 8. Monthly expenditure ratio (in %) for energy type (ET) to the total energy (TE) and to the total income (TI) for both biomass and non-renewables under the three income groups.
Category Energy type (ET)
Rich Medium Poor All respondents
ET: TE ET: TI ET: TE ET: TI ET: TE ET: TI ET: TE ET: TI Biomass
Firewood 44.75 4.61 49.75 8.62 26.99 6.61 41.94 5.91
Leaves and twigs
9.19 0.95 15.65 2.71 41.05 10.05 19.03 2.68
Bamboo 2.80 0.29 2.73 0.47 6.72 1.64 3.76 0.53
Crop residues 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Dung cake 0.00 0.00 3.08 0.53 12.31 3.01 3.96 0.56
Total 56.73 5.85 71.21 12.34 87.07 21.31 68.70 9.69
Non-renewable
Electricity 17.87 1.84 11.89 2.06 5.35 1.31 12.94 1.82
Kerosene 2.57 0.26 7.51 1.30 7.33 1.79 5.30 0.75
LPG 19.07 1.97 9.03 1.57 0.00 0.00 11.23 1.58
Candles 0.96 0.10 0.39 0.07 0.24 0.06 0.58 0.08
Natural gas 2.88 0.29 0.00 0.00 0.00 0.00 1.25 0.18
Total 43.35 4.46 28.82 5.00 12.92 3.16 31.30 4.41
Grand total 100.00 10.31 100.00 17.33 100.00 24.47 100.00 14.10
3.5 Energy expenditure and interactive factors
3.5.1 Energy expenditure ratio to total income of the households
Table 8 also illustrates the ratio of expenditure between energy type (ET) and total income (TI) within a month. The households, on average, spent 14% of their monthly income for energy
24
purposes, of which 10% accounted for biomass and 4% for non-renewables. Poor households spent much more of their income on biomass (21%), while the financially solvent households invested in non-renewables. This study reveals significant positive relationship (p < 0.01) between household income and total expenditure for energy consumption (Table 9). Household income was positively associated (p < 0.01) with the expenditure on firewood and non-renewable fuels (electricity, LPG, natural gas, candles), while being negatively associated (R= -0.20; p < 0.01) with that for leaves and twigs, and kerosene (Table 9). This indicates that households in a better financial condition are more likely to move from less convenient (e.g., kerosene) to more convenient fuels (electricity and LPG), confirming the finding of a recent study in South Asia, including Bangladesh, Nepal, and India by Behera et al. (2015). Table 8 also shows that poor households spent 21% of their monthly income on biomass, while rich ones spent almost the same amounts on both biomass (6%) and non- renewable energy (5%). The role of household income in the selection of fuels has been described in previous studies in south Asian countries, including Bangladesh, Nepal, and India (e.g. Akther et al., 2010b; Barnes et al., 2011; Behera et al. 2015; Bhatt and Sachan, 2004; Miah et al., 2011a, 2010; Van Ruijven et al., 2008).
Table 9. The association (Spearman ρ) between households’ energy expenditure and households’
size, income, landholding, and literacy.
Households’ energy expenditure Households' size Households’ income Landholding Literacy
Total energy 0.11 0.56 0.28 0.42
Total non-renewables 0.03 0.73 0.33 0.58
Total biomass 0.11 0.13 0.12 0.10
Electricity 0.11 0.75 0.31 0.61
Kerosene 0.03 -0.20 -0.03 -0.05
LPG -0.02 0.31 0.14 0.14
Candles -0.03 0.35 0.16 0.17
Natural gas -0.06 0.27 0.11 0.24
Firewood 0.11 0.47 0.29 0.39
Leaves and twigs 0.03 -0.43 -0.27 -0.36
Bamboo 0.09 -0.07 0.02 -0.02
Dung cake -0.00 -0.38 -0.16 -0.27
25
3.5.2 Energy expenditure, landholdings, literacy, and size of the households
Energy expenses were positively affected by the literacy rate of the household. A positive correlation (p < 0.01) between the households’ expenditure for firewood, electricity, LPG, natural gas, and literacy of the households’ members indicates the tendency of convenient fuel consumption with increasing literacy rate. Conversely, there was a negative relationship (R= -0.05; p < 0.01) between the households’ expenditure on leaves and twigs, dung cake, kerosene and literacy rates (Table 9). In line with our work, Miah et al. (2010) and Van Ruijven et al. (2008) also found that households with high literacy rates preferred spending more on firewood compared to other biomass fuels and using non-renewable fuels. Similar studies in Bhutan by Rahut et al. (2014), (2017) support our findings showing that Bhutanese households with higher education tend to switch from dirty fuels such as kerosene and dung cake to cleaner fuels such as electricity and LPG.
Thus, households’ literacy appears to be one of the key drivers of the energy switch from less convenient and dirty fuels to more convenient and clean fuels. This is also ascertained that households with more family members had to spend more for electricity and firewood use for lighting and cooking, respectively (Table 9), which was reflected in previous study (e.g. Hassan et al., 2012; Miah et al., 2011a; Rao and Reddy, 2007). Furthermore, the households’ energy expenditure and landholdings were positively correlated (p < 0.01), which indicate that the larger the landholdings, the higher the energy expenditure (Table 9). Consequently, the rich households with large landholdings were likely to spend large amounts of money compared to the medium- income and poor households with small landholdings (Fig. 2).
3.6 Carbon dioxide (CO2) emissions
The total emissions from the use of non-renewable energy accounted for 56.36 kg CO2 e household-
1 month-1,with the highest contribution from electricity. It should be noted that the burning of
26
firewood contributed the highest emissions (Table 10). This might be due to the traditional way of using biomass for cooking and heating. As illustrated in previous studies, the use of traditional cooking stoves generated higher emissions due to incomplete combustion, compared to improved cooking stoves (ICS) (e.g. Kurschner et al., 2009; Miah et al., 2009; Nath et al., 2013).
Table 10. Mean monthly emissions (kg CO2 e household-1 month-1)with standard error of mean (±) from using biomass and non-renewable energy in the study area.
Non-renewables Emissions (kg CO2 e household-1 month-1)
Biomass Emissions (kg CO2 e household-1 month-1)
Electricity 32.66 ± 2.29 Firewood 192.03 ± 10.57
Kerosene 3.82 ± 0.22 Dung cake 9.79 ± 1.54
LPG 5.42 ± 1.14 Crop residues 0.75 ± 0.43
Candles 0.16 ± 0.03
Natural gas 14.31 ± 4.41 Mean of total 56.36 ± 5.43
4. Concluding remarks and policy implications in rural Bangladesh
In Bangladesh, securing clean and equal energy access for all is often constrained by lack of understanding of households’ energy dependency and influencing factors. The major share of biomass energy, used primarily for cooking and heating at the household level, can be utilized in an efficient way to secure clean and equal energy access by reducing CO2 emissions from the combustion of biomass fuels. However, organized baseline data on energy resources, CO2 emissions from various energy sources, and relevant socioeconomic factors are lacking in the rural and suburban area. In this context, our research findings offer policy makers an insight into rural energy use patterns and enable them to develop tailored energy initiatives. This study explores and describes household-level energy consumption and expenditure of biomass and non-renewables and their comparisons, relevant socioeconomic factors and CO2 emissions from various energy sources.
Although this study has some drawbacks, including the overestimation of consumption and expenditure of energy fuels specifically for biomass by some households, it provides approximate findings verified by the researchers and local guides. In our study, most of the households heavily
27
depended on biomass energy, with the major fuels including firewood and leaves and twigs used mainly for cooking, and these were sourced mainly from their homesteads. About 24% of the poor who depended on public forests for biomass collection, 87% of them lived nearby the forests.
Regardless of the income groups, biomass energy accounted for households’ 87% of their monthly energy consumption and about two-thirds of energy expenditure and 10% expenses of total income.
Conversely, dependence on non-renewable energy was minimal and accounted for households’ 31%
monthly energy expenditure and 4% expenses of total income. The rich households tended to rely more on electricity, candle, LPG and their consumption of the non-renewables was significantly higher than that of medium-income and poor households. Both non-consumers and consumers of grid electricity, due to uneven rural electrification networks and power disruptions, mainly used kerosene and candles for lighting. The medium-income and poor households had to spend significantly more of their income on kerosene. Income, education, and landholdings of households are positively correlated with expenditure for consuming convenient energy sources such as firewood, electricity and LPG. However, firewood as a cooking fuel was an energy source with the highest carbon emissions (192 kg CO2 e household-1 month-1).Regardless of the income groups, the households’ consumption of biomass was 555% higher compared to that of non-renewables. Hossen et al. (2017) also stated that efficient use of only one-third of the overall biomass available would meet the total energy demand in Bangladesh, thereby avoiding the use of fossil fuels.
The Renewable Energy Policy 2008 of Bangladesh aims to harness the potential and dissemination of renewable energy resources and their technologies, for example, biomass gasification and clean energy promotion for CDM, while substituting the non-renewable energy resources (GOB PD, 2011). Our results can relate the renewable energy policy; as such, the biomass consumption of 1087.79 kWh could be alternatively used by the installation of biomass-based small-scale power plants and gasification systems at the village level to produce electricity and heat to fulfil the local demand of energy. This clean energy use would result in less carbon emissions
28
compared to the direct combustion of biomass, and simultaneously would reduce the dependency on the non-renewable fossil fuels through substitution. Cooking fuel conversion efficiency is very low (Mainali et al., 2014), which led to the overuse of biomass collected from homestead forests. Under the CDM activities, introducing improved cooking stoves (ICS) to rural households can substitute natural gas and LPG for cooking, and can concurrently reduce overdependence on the homestead and public forests, thereby preventing overexploitation (Miah et al., 2009, 2011b; Nath et al., 2013).
The initiation of CDM forestry programs in homesteads and cropland agroforestry systems could ensure the sustainable supply of biomass while maintaining environmental sustainability and offering an alternative livelihood option for poor households through carbon trading (Miah et al., 2011b). Homestead agroforestry is believed to have a higher potential to sequester carbon than pastures or field crops and, consequently, to add higher carbon credits (Nair et al., 2009; Nath et al., 2014). Uddin and Taplin (2009) also found CDM being potential in sustainable energy projects in Bangladesh. However, lack of an assigned government body for promoting biomass energy may hinder the CDM forestry project, which also needs coordination between local land users and different government organizations, such as forest department and energy department. The opportunity costs of agriculture and homestead production may influence the decision of CDM project for biomass production and carbon sequestration (Smith, 2002).
In our research, we did not study the use of ICS and feasibility of biomass-based technologies under the sustainable energy projects. Therefore, policy supports, together with research and design (R&D) on the feasibility of biomass-based technologies and accessibility to ICS at the village level are essential prior to the widespread adaptation of such technologies. The GOB should invite local land users, research institutions, and international donor agencies/NGOs to work together with concerned government organizations on technological development and biomass-based energy to secure clean and equal energy access for all. Additionally, augmenting the income of the rural households through alternative livelihoods, such as carbon trading, while increasing the education
