2.1. Case Study Area: Geographic, Agro-Ecological, and Socio-Demographic Characteristics
The state of West Bengal was selected as case study area because it is one of India’s major food baskets, with 70% of the population directly engaged in agriculture [
21]. Of these, 90% are smallholder farmers [
22]. A quarter of the population has no access to electricity, while 40% partly has access, with supplies however being limited and unreliable [
8]. Agro-ecological conditions are favorable: the area is endowed with strong irradiance, fertile alluvial soils, and plentiful water resources [
23]. Average precipitation is 1435 mm a
−1 [
24]. In the dry season, channels from the Hooghly River and groundwater are used for irrigation. Abundant labor resources and access to agricultural inputs (fertilizers, pesticides, high-yielding varieties) result in high crop productivity. Land availability is the major constraint for agriculture [
23]. Today, virtually all of the land is utilized and there are very few natural forests [
25,
26]. West Bengal is a traditional rice-growing area, but since 1997 there has been a shift towards the cultivation of fruit, vegetables, and flowers, as these provide higher revenues [
23]. Based on these local characteristics, the area provides a suitable setting to assess the potential of smallholder RE production. Research was conducted from January to March 2015 in the two rural villages, Ghoragachha and Baikunthapur, in collaboration with the Department of Agricultural Extension of the State Agricultural University Bidhan Chandra Krishi Vishwavidyalaya (BCKV, Mohanpur, WB, India) under the Indian Council of Agricultural Research.
2.1.1. Ghoragachha
Location: Chakdaha Block of the Nadia District (88.31° E; 22.58° N).
Population: 1830 inhabitants in 134 households. The population (100% Muslim) is defined as “minority” in the official classification of the Indian population. The vast majority are smallholder farmers. Only 44 inhabitants are engaged in other jobs (e.g., embroidery, government sector). All households are connected to the grid [
25].
Area size: 2825 acres (11.43 km
2), about 2000 acres agricultural land, 20 acres water bodies used for fish farming, the remaining area is residential [
25].
Access to the village was facilitated through the Department of Agricultural Extension (BCKV, Mohanpur, WB, India), frequently organizing workshops about the introduction of new farming practices and technologies (fertilizers, pesticides, machinery). Thus, farmers were familiar with household interviews and technology evaluation. They willingly shared information about their livelihoods and perceptions regarding RET often engaging in open and critical discussions.
2.1.2. Baikunthapur
Location: Amdanga Block of the North Twenty Four Parganas District (88.32° E; 22.55° N), about 25 km south of Ghoragachha.
Population: 2940 inhabitants in 263 households, mainly belonging to the Hindu caste of Vaishyas. Most farmers have additional sources of income besides agriculture. Along the main road, there are 31 handicraft workshops, mainly carpentry and metal-work, and also a few agro-processing and Sheetal Pati (cool mat) weaving enterprises [
26]. The village is connected to the grid [
26], except for the poorest farming households on the outskirts of the village, where research was conducted.
Area size: approx. 700 acres (2.83 km
2), 550 acres farmland, 6 acres lakes, and residential area on the remaining land [
26].
The additional interviews at Baikunthapur showed that the activities of the Department of Agricultural Extension from BCKV have not influenced the perceptions of the smallholders in Ghoragachha towards RE production. The Baikunthapur farmers willingly shared information about their livelihoods. The farmer livelihoods in both villages are very similar.
2.2. Integrated Renewable Energy Potential Assessment (IREPA)
IREPA is a participatory, bottom-up approach to REP assessment and appropriate RET selection in smallholder farming systems of developing and emerging economies. It integrates the biophysical perspective of natural science (RR assessment), the socio-cultural perspective of social science (smallholders’ livelihoods), and the techno-economic perspective of energy engineering (RET planning) in an inter- and transdisciplinary approach. IREPA was previously tested in a rural area of South Africa [
20] and adapted to this specific context, applying the following steps:
Access to the two villages was facilitated by the Department of Agricultural Extension of the BCKV. Further, it was possible to draw on geographic, agro-ecological, and socio-demographic profiles and statistics from the Directorate of Micro, Small, and Medium Enterprises of the Government of West Bengal [
25,
26].
The bio-physical RR availability of solar, wind, hydro, geothermal, and bioenergy was assessed according to the REP categories defined by [
9,
27]. The RR data were obtained from international and national databases and scientific literature (
Table 1).
Multiple participatory research methods [
33] were applied to explore the household characteristics and the structure of the two villages (
Table 2). Details of local livelihoods were obtained from semi-structured, open-ended household interviews (
n = 29) with the following question categories, drawing on the structure of the Sustainable Livelihoods Framework (SLF) [
34]:
Human Capital: education level, jobs and sources of income, other skills, work distribution, openness to innovation;
Social Capital: household members, age, decision making, water/sanitation, health, food sources and preparation, gender roles;
Institutional: social assistance, agriculture, and energy subsidies;
Natural Capital: environmental concerns, access to natural resources, details of the agricultural system including crops, animals, and inputs;
Physical Capital: land ownership, waste disposal, energy supply and use patterns, knowledge of RE.
Interviews, transect walks, and participant observations were always accompanied by one of the three students from BCKV, who translated from Bengali into English and vice versa. The information obtained was transcribed and coded to allow categorization, generalization, and interpretation [
35] (p. 184ff). Subsequently, the information was cross-checked and triangulated with data from secondary sources listed in
Table 2 and the key informants from BCKV. Quantitative data were processed using MSO Excel 2007, performing descriptive statistics to calculate percentages, averages, quantiles and ranges (min–max) [
35].
The results were structured using the SLF, characterizing people’s livelihoods based on five categories of Livelihood Assets (human, social, physical, natural, and financial). These are influenced by the external environment, which in turn affects the wider availability of assets, and circumstances people have to cope with, such as natural disasters, financial crises, seasonality, and others (Vulnerability Context); institutions, organizations, policies, and legislation (Transforming Structures and Processes); and the strategies adopted to achieve certain Livelihood Outcomes (Livelihood Strategies) [
34].
These results were then categorized according to relevant social, institutional, environmental, technical, and economic factors, regarded as local RET sustainability criteria. Eight criteria were derived that determine the requirements for an RET to be appropriate in this specific local context. Not more than nine criteria should be employed in the Analytical Hierarchy Process (AHP) (Step 6) to keep execution manageable and increase consistency of results [
32].
Biomass resources available for energy production (and cultivation areas) were explored at individual farm level. Local yield levels were obtained through household interviews supplemented by FAOSTAT data [
36].
The GREP determines the suitability of land areas for RE generation based on bio-physical RR data including temporal patterns [
9,
27]. Local RR power densities were evaluated based on threshold values for technical utilization. Average solar irradiance at ground level should be at least 120 W m
−2 [
27]. For wind energy, the average wind speed needs to exceed 4 m s
−1 at 10 m hub height [
7]. Power densities below these threshold values were excluded from the assessment. For hydro energy, the proximity of a river and its accessibility for smallholders was evaluated. For geothermal energy, the availability of easily accessible geothermal sources (e.g., hot springs) was assessed. The type, amount, and availability of the biomass resources obtained in IREPA Step 3 were analyzed. Only biomass residues or wastes without alternative uses, such as animal feed, were considered for energy production. The potential bioenergy output was estimated based on average values of residue-to-product ratios, energy content and biogas yields obtained from the literature (see
Section 3.2.5).
The availability of suitable areas (and biomass residues and wastes) for RET installation was evaluated based on the local RET sustainability criteria (IREPA Step 3). Then, the conversion efficiencies of locally available RET were factored in [
9,
27].
Locally appropriate RET were pre-selected based on the TREP and the local RET sustainability criteria (IREPA Step 3), applying a cradle-to-grave value chain perspective for RE generation.
Four RETs were proposed and planned by the researcher, taking available materials and technology devices into account, including their costs and operation and maintenance (O&M) requirements.
Local appropriateness of the proposed RETs was assessed by applying two multi-criteria decision analysis (MCDA) methods with local participants: an adapted Analytical Hierarchy Process (aAHP) [
37] and the Simple Multi-Attribute Rating Technique (SMART) [
38]. The local RET sustainability criteria served as selection criteria in both MCDA methods.
The AHP is the most frequently used MCDA method in sustainable energy decision-making [
12]. This descriptive method allows stakeholders (e.g., smallholder farmers) to compare and rank a set of selection criteria and RET options, arranged in pairs, to identify the most appropriate technology. However, the application of the AHP was found to be time-consuming and in some cases too abstract for the smallholders interviewed in the South African IREPA case study [
20]. Therefore, the standard AHP [
12] was adapted by clustering the RET sustainability criteria based on qualitative similarities [
39] into:
(i) Overall benefits associated with RET implementation; (ii) Benefits associated with RET; (iii) Energy-use benefits; and
(iv) Requirements for RET implementation. Criteria clustering reduces the number of pairwise comparisons and helps participants clarify their thoughts, while increasing feasibility of application and accuracy of results [
39]. Seven of the farmers previously interviewed participated in the aAHP questionnaire. First, the local RET sustainability criteria were arranged in pairs. The farmers rated the importance of the two criteria against each other on a scale of 1 to 9 (1 = equal importance; 9 = absolute importance over the other). Second, each participant compared the proposed RETs, arranged in pairs, with each other, and with respect to each sustainability criterion. The pairwise comparisons allow the computation of a decision matrix. Based on these ratings, the eigenvector of (i) each sustainability criterion and (ii) each RET are calculated using common matrix algebra. This expresses the priority of each participant [
37]. However, the farmers found the abstract nature of pairwise comparisons difficult and complained about the time-consuming procedure.
For this reason, SMART was applied with five previously interviewed key informants instead of aAHP. SMART is based on a straightforward ranking of criteria and RET. First, the RET sustainability criteria were rated on a scale of 1–100, assigning the most important criterion the highest score, the second-most important criterion the second-highest score, and so on. Then, the four RETs were evaluated based on their performance/compliance with each RET sustainability criterion, assigning the highest score (1–100) to the best-performing/complying RET, the second highest score to the second best-performing/complying RET, and so on. Thus SMART is easier and less time-consuming to apply, as scores are directly assigned to each criterion and to each RET option [
38]. Finally, the rankings obtained from aAHP and SMART revealed the most appropriate RET, at village level (taking the average results of all rankings) or at household level (taking the ranking of an individual farmer).
Differentiation between farmers’ and key informants’ rankings allowed a comparison of viewpoints based on personal situation between insiders (farmers) and outsiders (key informants).
Once IREPA has been applied, all stakeholders concerned (local smallholders, government bodies, companies and non-governmental organizations) should have comprehensive information at hand. If an RE project is to be realized, these stakeholders should be able to select locally appropriate RET with a high chance of successful implementation.