A Comparative Study Of Small-scale Rural Energy Service Pathways For

Transcript

Applied Energy 101 (2013) 376–392 Contents lists available at SciVerse ScienceDirect Applied Energy journal homepage: www.elsevier.com/locate/apenergy A comparative study of small-scale rural energy service pathways for lighting, cooking and mechanical power Mirco Gaul ⇑ Institute for Energy Engineering and Centre for Technology and Society, Technische Universitt Berlin, Sekr. HBS 1, Hardenbergstr. 16-18, 10623 Berlin, Germany h i g h l i g h t s " Life cycle based analytical framework for rural energy service pathways (RESPs). " Comparison of Jatropha-based RESPs for lighting, cooking and mechanical power. " Weak performance of Jatropha oil in the categories lighting and cooking. " The potential for power depends on capital, energy, labour, and transport intensity. " Simultaneous use of plant oil and biogas at village scale shows best potential. a r t i c l e i n f o Article history: Received 13 July 2011 Received in revised form 9 March 2012 Accepted 24 March 2012 Available online 23 May 2012 Keywords: Rural energy service pathway Energy analysis Cost analysis Life cycle assessment Jatropha curcas Indonesia a b s t r a c t The strong international growth of biofuels in the last decade brought the interest in bioenergy back on the agenda. While many life cycle assessments for biofuels mainly focus on environmental impacts and costs, over the last decade especially the energy balance of biofuel production chains has been a major point of criticism. This study applies a specially adapted and LCA-based analytical framework for rural energy service pathways (RESPs) to compare the use of Jatropha plant oil and biogas with other small-scale RESPs for lighting, cooking and mechanical power. The aim is to analyse their technological feasibility and economical viability by comparing the energy and cost efficiency. Results show strong differences for the investigated plant oil production and processing pathways, while the comparison with a baseline and a competitive renewable energy scenario reveals a weak performance of plant oil and even biogas in the categories of lighting and cooking. The potential for mechanical power depends largely on the careful optimisation of the energy service pathway by balancing the capital, energy, labour, and transport intensity. For the present case, the village scale production of Jatropha plant oil and biogas and their simultaneous use in a dual fuel engine to locally provide power and electricity would be the service pathway with the highest potential in terms of energy and cost efficiency. Ó 2012 Elsevier Ltd. All rights reserved. 1. Introduction Service delivery in remote rural areas of developing countries remains a major challenge for both, electrical and non-electrical energy services. Latest data confirm the slow decrease of the number of people without access to electricity from 1.6 billion in 2002 to 1.4 billion today and an expected 1.2 billion by 2030 while the number of people relying on the traditional use of biomass for cooking and heating is still increasing together with the global population from 2.4 billion in 2002 to 2.7 billion today and an expected 2.8 billion by 2030 [1,2]. In any case, there remains a substantial gap between the objective to reach universal energy access by 2030 [3] and the business as usual scenario as described ⇑ Tel.: +49 176 27457740. E-mail address: [email protected] 0306-2619/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.apenergy.2012.03.050 by the IEA. Due to the population increase, particularly in developing countries, the share of bioenergy in the continuously growing global energy demand remained stable at 10% in the past decade. Of this global bioenergy consumption, the share of so-called ‘modern bioenergy and biofuels’ is only 22%, while the remaining 78% comprise traditional use in rural areas where bioenergy often makes up over 90% of the total energy demand [4,5]. These rural areas are typically characterised by a weak infrastructural setup in terms of healthcare, education, sanitation and transportation. Alongside firewood and dung, kerosene and candles are traditional fuels for cooking and lighting, whereas liquefied petroleum gas (LPG) and electricity often are unavailable or cost-prohibitive for most of the rural population. The burden of firewood and dung collection for cooking rests mainly on women and female children, who at the same time are most affected by the indoor air pollution caused by inefficient stoves and open cooking fires. Worldwide, M. Gaul / Applied Energy 101 (2013) 376–392 377 Nomenclature Abbreviations GHG Green house gases ICT information and communication technologies IDR Indonesian rupiah, with 12,819 IDR/€in February 2010 IEA International Energy Agency LCA life cycle assessment LED light emitting diode LPG liquefied petroleum gas MHP micro-hydropower RESP rural energy service pathway RET renewable energy technology SHS solar home system EPri EUse EUsea EUset GER n NEB NEC NER NEV NPV ROL gUse Parameters ALD annual labour demand (wd/a) ANECBase annual net energy costs of the baseline (€/a) ANECExL annual net energy costs without labour (€/a) total benefits of the system in year t Bt Ct total costs of the system in year t CCap annual capital costs (€/a) annual fuel costs (€/a) CFuel CInp annual input costs (€/a) CInv initial investment costs (€) CLab annual labour costs (€/a) annual transport costs (€/a) CTra CED cumulated energy demand (kW h/kW h or kW h/kLm h) d Real discount rate (corrected for inflation) (%) EAux total auxiliary energy input (kW h) supplied final energy to the end-use process (kW h) EFin almost two million deaths annually from pneumonia, chronic lung disease, and lung cancer are associated with exposure to indoor air pollution resulting from cooking with biomass and coal. Of these deaths, 99% occur in developing countries [6]. As it is widely accepted that even for the coming decades large parts of rural population will not be connected to central power grids or fossil fuel logistic chains [2,7], the development of decentralised rural energy systems using local renewable resources has come into greater focus. Renewable energy technologies (RETs), increasingly employed in the rural context, include small and micro-hydropower turbines, solar home systems (SHSs), biogas digesters, and improved stoves [8]. Biomass remains the primary energy source for cooking and heating in rural as well as many urban areas. Due to growing populations and the subsequent rise in land, water and energy demand, urban and rural settlements and their energy systems have in many areas increased the pressure on local resources. While natural firewood resources diminish with growing speed and the prices for fossil fuels climb rapidly, the regions concerned are now faced with varying degrees of environmental stress. Although wood clearance for new agricultural land (and not for firewood) is in many areas the driving factor for deforestation, soil erosion and desertification, the growing firewood demand can often not be fulfilled with the shrinking wood resources [9]. The ongoing debate on how to improve traditional biomass use and strategically develop bioenergy potential in developing countries [10,11] has been further sharpened in the last decade when the demand for bioenergy and biomass increased globally, as a consequence of public policies of industrialised countries in the context of climate change and energy security [12]. Biomass resources can theoretically be used in many different conversion pathways to cater to all kinds of (rural) energy services, Units a h J l Lm m W wd total primary energy input (kW h) levelised useful energy output (1 kW h or 1 kLm h) average annual useful energy output useful energy output in year t Gross Energy Ratio (kW h/kW h or kW h/kLm h) lifetime of the project (a) Net Energy Balance (kW h or kLm h) Net Energy Cost (€/kW h) Net Energy Ratio (kW h/kW h or kW h/kLm h) net energy value (kW h or kLm h) net present value (€) return on labour (€/wd) efficiency of the end-use appliance year, unit of time, 8760 h hour, unit of time Joule, unit of energy litre, unit of volume Lumen, unit of the luminous flux metre, unit of length watt, unit of power workday, unit of the labour demand (assumed as 8 h per day) SI-Prefixes k kilo, 103 M mega, 106 but not all of these pathways are equally favourable in terms of resource, conversion or cost efficiency and green house gas (GHG) emissions. For the present research, a project using the oil of the tropical scrub Jatropha curcas L. has been chosen since Jatropha oil provides the above stated flexibility and the Jatropha System has been discussed as a particularly promising example of small scale bioenergy development. The expression Jatropha System has been introduced, among others by Henning [13] based on experiences in the early 1990s in Mali and describes the smallscale cropping, processing of Jatropha for local energy demand and income generation.1 Globally, the bioenergy debate is focusing on the supply side, analysing cost, accessibility and competing use of agriculture residues, as well as agricultural performance and land use impacts of specific ‘energy crops’ [5,22,23]. In the case of Jatropha research concentrates predominantly on plant characteristics and cropping systems, or the possible use of Jatropha oil for biodiesel production [24–27]. Little is known about the potential of Jatrophabased rural energy services. This is due to the limited practical experience and research on Jatropha to date as well as due to the complexity and difficulties of comprehensive energy analyses in 1 Even though Jatropha has already been planted as an oil producing scrub a century ago, systematic research has just started during the last two decades and gained momentum only in 2007. By the end of 2011, WorldCat lists a number of more than 2200 books and journal articles on J. curcas, of which only 17% were published before 2007 [14]. Heller [15], Gübitz [16] and Jongschaap [17] provide overviews on plant characteristics, cropping and processing, while GEXSI presented 2008 a very optimistic global outlook on Jatropha projects and activities [18]. The FACT Jatropha Handbook [19] and an IFAD/FAO publication [20] collect many practical information on small-scale Jatropha systems, however do not provide an overview on scientific research. To provide this overview is the aim of a study commissioned by the NL Agency of December 2010 that evaluated 200 studies on agronomic and socioeconomic aspects of Jatropha [21]. 378 M. Gaul / Applied Energy 101 (2013) 376–392 Table 1 End-use oriented analytical framework of energy supply pathways. Scenario Cooking Lighting Mechanical power Baseline Jatropha Competitive RET Existing stoves Plant oil & biogas cookers Solar & Improved wood stoves Existing lamps Plant oil & biogas lamps Solar lamps Existing engines Plant oil, biodiesel & biogas engines Micro hydro turbine general that also characterises the ongoing debate on the feasibility of large-scale biofuel production. Most energy system analyses follow a life cycle assessment (LCA) approach2, but because of different system boundaries, functional units and allocation rules, LCA results often are contradictory and difficult to compare. This is particularly true for LCA of bioenergy systems with very complex process chains [32–35]. Current efforts therefore try to increase the transparency of assumptions and parameters [36] and to agree on sustainability indicators [37]. Another reason for differences in analysis results are the selected analysis parameters. For the energy analysis of biofuels often the net energy balance or net energy gain is used, which is defined as the difference of the energy output and energy inputs of biofuel production [38–41]. However, often only the fossil share of energy inputs is counted, an approach that has been criticised as not adequate to evaluate the feasibility of biofuel production, as the energy efficiency of the conversion process is not clearly described [42]. The present research analyses the technical and financial feasibility of complete small-scale Jatropha Systems for the local energy demand in rural areas. Therefore a specially adapted and structured analytical framework for rural energy service pathways is applied to the case study of a Jatropha project that is situated on the Indonesian island of Sumbawa. Interviews were conducted during a project visit in February 2010 both with experts of government and private sector as well as with farmer groups in 10 villages that grow and process Jatropha seeds. The aim of the research is to analyse and compare the Jatropha-based service pathways with existing or realistic alternatives. To that end, first a baseline scenario and a competitive RET scenario are defined as benchmarks. Then three different Jatropha scenarios were developed, focusing on the household, village or region as scale and production unit. All five scenarios were evaluated and compared regarding their energy and cost efficiencies. 2. Method The present analysis follows a life cycle approach and uses the life cycle analysis programme and database GEMIS 4.6 [43]. According to the ISO 14040 terminology a life cycle inventory analysis is conducted in the present study focusing on energy and costs without conducting a full life cycle impact assessment. Subsequently the approach and proceeding will be presented in detail. 2.1. Goal, scope and system borders The analysis follows a service perspective based on the assumption that despite the high complexity and diversity of rural settlements, the energy service demand and the related end-use in remote rural areas follows a clear pattern of four main service categories: (1) lighting, (2) cooking, (3) mechanical power for machines and transport, and (4) electricity for ICT and small 2 LCA is a method that has a long tradition for analysing product chains as well as energy systems [28,29]. International standards for LCA have been established in 1997 as ISO 14040 and ISO 14041-43, which have been revised in 2006 merging the latter into ISO 14044 [30,31]. The norm defines important key terms and the four LCA phases, namely: (1) goal and scope definition, (2) inventory analysis, (3) impact assessment and (4) interpretation. appliances. Because bioenergy can only be converted into electricity via the intermediate conversion into mechanical power (by a steam or internal combustion engine), the last service category is not further explored in the present analysis. To evaluate the efficiency of bioenergy use a comparative approach is followed, which includes a baseline scenario and a competitive RET scenario. Efficiency and cost indicators are levelised to the useful energy output, which is, in the case of lighting, the luminous flux; in the case of cooking the heat that enters the cooking pot; and in the case of mechanical power, the power on the shaft. Table 1 provides an overview on the selected end-use technologies of the analysis. Most LCA for biofuels and Jatropha focus on the production of liquid fuels (as final energy carrier) and the related land use and GHG emissions [41,44,45], while the present analysis focuses on the full rural energy service pathway that also includes the enduse appliance that provides the useful energy required by the energy service. The energy and cost performance of the different RESPs to provide useful energy in the three categories of lighting, cooking and mechanical power are compared. The option to locally produce and distribute electricity via a mini-grid has not been included. Regardless of whether electricity is generated by a diesel, plant oil or biogas engine powered generator or a hydropower turbine, the additional cost and energetic losses for generator and distribution will be similar, while the difference of the service chain is mainly caused by the upstream processes that are covered by the category mechanical power. For the analysis the RESP is divided into the four sub-systems: extraction, conversion, distribution, and end-use as shown in Fig. 1. The first sub-system ‘extraction’ covers the cropping and harvest of Jatropha fruits. It is only analysed by the required auxiliary energy and cost inputs, and by the provided primary energy output (in the form of Jatropha fruits). The natural resource inputs are not included into the calculation (illustrated by the doted line in the figure that highlights the balance area). This is justified by the cropping system of the selected case study, in which Jatropha is planted only as a living fence around existing fields and the practice of living fences has already been an established practice in the project region. Due to these circumstances no additional inputs for the establishment of the Jatropha fences are required and no land use competition occurs. The selected case study consequently represents the optimum case regarding a smallholder cropping practice of Jatropha-based RESPs and the present analysis evaluates the potential performance under such optimal circumstances. The system includes the energy end-use in the three categories of lighting cooking and mechanical power, because typical end-use appliances in rural areas have very low efficiencies and therefore significantly influence the overall system performance. The functional unit for this analysis is consequently the provided unit of useful energy defined as one kilowatt hour (kW h) for cooking energy and mechanical power and one kilolumen3 hour (kLm h) for lighting. The energy service itself (e.g. an illuminated room or cooked food) is not covered by this quantitative analysis as other case specific parameters influence the quality of the energy service (e.g. 3 Traditionally, the power of a light bulb has been specified using the unit watt to describe the total radiant flux (light and heat combined) but not its performance as a light source. Recently, the luminous flux (measured by the unit lumen) is increasingly used as it describes only the visible share of the radiation of a light source. 379 M. Gaul / Applied Energy 101 (2013) 376–392 Fig. 1. Definition of a RESP and the applied balance area. the size and surface colour of the room, or cooking habits and the type of cooked food). 2.2. Energy analysis The present analysis uses four parameters to characterise the energy system. In a first step, the cumulated fossil and non-fossil energy demand (CED) for the provision of one unit useful energy is calculated [46]. In Fig. 1, the CED is divided into the primary energy input EPri and the required amount of auxiliary energy EAux, typically in the form of electricity or fuels. In a second step, building on the levelised unit of useful energy output EUse, the Gross (GER) and Net Energy Ratio (NER) as well as the Net Energy Balance (NEB) are defined as shown in Fig. 1 and formula (1)–(3). EUse EUse ¼ CED EPri þ EAux NEB ¼ ðEFin  EAux Þ  gUse GER ¼ NEB ðEFin  EAux Þ  gUse NER ¼ ¼ EPri EPri ð1Þ ð2Þ following years. Consequently, the time value of the costs5 is used for the economic analysis of the bioenergy conversion pathway defining net energy costs (NECs) by the ratio of the net present value (NPV) of the bioenergy system costs and the useful energy supplied [48]. For a comparative evaluation, the energy costs of the baseline are first calculated (setting Bt = 0) and expressed as a positive value and then considered as saving and therefore included as benefit in the calculation of the comparative scenarios (formula (4)). Pn Bt C t t¼0 ð1þdÞt NPV ¼ Pn Euse NEC ¼ t NEV t In the present case study the whole investment is done in the year zero and both annual costs and supplied useful energy per year can be assumed to be stable over the project life time. In this case the NEC can be directly calculated by dividing the capital, fuel, transport, input, and labour costs by the annually supplied useful energy (formula (5) and (6)): ð3Þ The GER is the ratio of the energy output versus the total (both fossil and non-fossil) energy input. It is used to describe the efficiency of a conversion process, but does not reflect the different qualities of the primary energy input and the auxiliary energy. The NEB provides the difference of energy output and auxiliary energy and therefore reflects the fact that the auxiliary energy needs to be provided additionally (again causing losses in the conversion from primary energy). Because the compared systems provide different types of usefull energy, for the calculation of the NEB the auxiliary energy EAux is subtracted from the final energy EFin instead of the useful energy output. The difference is then multiplied by the efficiency of the end use appliance gUse (see formula (2)). As a combination of both approaches, the NER can be calculated as a ratio of NEB to energy input and reflects the real energy efficiency of the conversion process. A particular interpretation of the definition of energy ratios by useful energy output instead of final energy output occurs in the case of lighting: because the energy ratio is expressed in kLm h/kW h rather in per cent, the energy ratio value can be larger than one.4 ð4Þ t¼0 ð1þdÞ C Cap þ C Fuel þ C Tra þ C Inp þ C Lab EUsea ð1 þ dÞn ¼ C Inv  d ð1 þ dÞn  1 NEC ¼ ð5Þ C Cap ð6Þ In rural areas, one of the most controversial costs are labour costs for a population that lives from subsistence farming, some cash crops, and occasional daily labour. At the same time labour cost represents the largest cost-share in the case of labour-intensive process chains. For the present analysis typical local salaries for rural daily workers have been identified in interviews to be approx €1.95 per workday (wd) and cross-checked with statistical data on the rural poverty line of Sumbawa Island (about €12,83 per person and month in 2009) [49]. Considering the average household size for rural Sumbawa of 3.95 person and only one family member working, the required daily income (with 25 working days per month) would be €2.03 just to reach the poverty line6. The assumed labour cost of €1.95 per man-day therefore seems already in the lower range and a calculation with lower values would risk a negative development impact. However, as an alternative, the return on labour (ROL) of an energy service pathway is introduced [48] as the difference of the an- 2.3. Cost analysis Like most energy systems, bioenergy systems require a substantial initial investment while the benefits only accumulate over the 4 This is caused by the fact that the theoretical optimum of the luminous efficiency is about 683 Lm/W, while values as 200 Lm/W have been already reached in laboratory and the selected LED lamp for the present analysis has a efficiency of 100 Lm/W. 5 Calculating net present values does not automatically mean the application of a comprehensive cost-benefit analysis (CBA). Such an analysis would not make sense in a small-scale environment, as CBA methods are typically applied to decision-making at national level. For a current discussion of CBA methodology see [47]. 6 The current data for rural Sumbawa in 2010 has not been available, but if the general increase of the rural poverty line of Indonesia of about 7% compared to 2009 [50] is also applied to Sumbawa, the required income per man-day would increase to €2.17. 380 M. Gaul / Applied Energy 101 (2013) 376–392 2.4. Comparative scenario development Fig. 2. Baseline scenario. nual net energy costs without labour costs (ANECExL) to the annual net energy costs of the baseline scenario (ANECBase) and divided by the annual labour demand (ALD) in workdays as shown in formula (7). ROL ¼ ANECBase  ANECExL ALD ð7Þ The ROL shows in how far the savings of the bioenergy service pathway in comparison to the energy baseline system justify the labour investment for the rural farmer. In the case of cooking, the baseline 3-stone-fire has no monetary costs and for the calculation of the ROL the cost of a kerosene wick stove was used as a baseline. Thus the ROL for cooking pathways has to be interpreted with care, as the majority of the rural population does not use kerosene stoves and cannot profit from such high savings. Subsequently the analysed scenarios are briefly described, while the specific input data is presented in Tables 4–7. The baseline scenario for rural areas of Sumbawa has been developed based on local research and interviews with farmers’ groups in 10 villages and is shown in Fig. 2. It consists of four energy service pathways: (1) the provision of firewood for cooking on a 3-stone fire, the provision of kerosene fuel for cooking and lighting by (2) a kerosene stove and (3) a kerosene lamp and (4) the provision of diesel fuel for mechanical power by a diesel engine. The diesel and kerosene supply chains are mainly external, consisting of crude oil extraction and conversion on the Indonesian main island Java and the distribution via tank ships to the two Pertamina (the Indonesian state oil and gas utility) fuel depots on Sumbawa and further via 15 filling stations and 8 kerosene distributors on the island. For the calculation of the fossil fuel chain, existing GEMIS data sets for India [51] and Indonesia have been adapted to the specific context [52]. For the cost calculation only the final local prices for diesel and kerosene have been considered as the cost structure of the highly subsidised fossil fuel supply in Indonesia is difficult to assess [53]. Critical aspects of the supply pathway are the informal transport and retail structures that supply rural villages at a sharp increase of selling prices and the low efficiencies of rural end-use technologies. The supply chain for firewood-based cooking is fully local as in all visited villages wood is locally collected from nearby sources. Even if the Jatropha energy system reveals itself to be advantageous compared to the baseline scenario there might still be better alternatives. Therefore, for this analysis four of the most promising renewable energy technologies alternative have been selected as competitive RET scenario as shown in Fig. 3: (1) for cooking a solar stove; and – because the feasibility and acceptance of solar stoves has not yet been demonstrated for Sumbawa – also (2) an improved wood stove; (3) for lighting a Solar LED lamp,7 as they represent the current most cost efficient way of rural electric lighting for low lighting intensities; and (4) for mechanical power a microhydropower (MHP) turbine, because significant experiences on MHP already exist in Indonesia.8 All three Jatropha scenarios share the same extraction or cultivation sub-system based on the cultivation of Jatropha scrubs as living fences to protect the fields from grazing animals. Because fences are built anyway by the farmers and no fertilisers, pesticides or other inputs are applied, no costs are allocated to this production system. However, the picking and hulling of fruit is very time-consuming. The household scale Jatropha scenario is shown in Fig. 4. For the conversion sub-system, manual cleaning is compared with the use of a simple mechanical hulling tool, defining the two sub-cases H1 and H2. After the cleaning there are two alternative conversion pathways as shown with the continuous and dotted arrows. The continuous arrow shows the pressing of Jatropha seeds with a manual ram-press, and the subsequent direct use of the plant oil by the farmer for cooking and lighting, using a plant oil wick stove or a wick lamp. The dotted arrow shows an alternative pathway where the farmer uses the press-cake to produce biogas with a family size biogas digester as local fuel for cooking and lighting using a biogas-adapted LPG stove or lamp. The plant oil is sold as a byproduct via a local coordinator to a biodiesel factory, and the factory sells the biodiesel to the regional fuel de- 7 As a current example the Solux LED 105 has been selected, which uses a CREE Q5 LED with approx 100 Lm/W efficiency and has a price of €40 in bulk that has been increased to €60 to reflect import and retailing costs [54]. 8 Even though on Sumbawa only one MHP scheme has been developed so far by the government, more than 100 MHP schemes have been built on other islands and average data is based on that experiences (see Table 6). 381 M. Gaul / Applied Energy 101 (2013) 376–392 Fig. 3. Competetive RET scenario. Fig. 4. Jatropha scenario – household scale. pot, from which it enters the established distribution pathway for fossil fuels and in this way is also available to the farmer as engine fuel. The heating value of the byproduct plant oil is deducted from the CED of the produced biogas and the profit from selling the plant oil is allocated to the biogas system as negative costs. The press-cake or the slurry from the digester can be used as fertiliser in the fields, but is not allocated as negative cost. Thus, the household based scenario defines ten service pathways: plant oil stove H1 and H2, biogas stove H1 and H2, plant oil lamp H1 and H2, biogas lamp H1 and H2, and biodiesel engine H1 and H2. The village scale Jatropha scenario is highlighted in Fig. 5. For the conversion sub-system, the option of mechanical hulling by the farmer is compared to a engine-powered hulling machine operated by the local coordinator in the village. In this way the two subcases V1 and V2 are defined, of which only the second is shown in Fig. 5. In case V2 the farmer has to transport the whole fruit to the village for hulling. After hulling the seeds are pressed with an engine-powered screw press in the village. Two alternative pathway options are marked by the continuous and dotted arrows. The first option is to directly use the plant oil in the village by distributing it as cooking and lighting fuel or to run a plant oil engine for mechanical power or electricity generation in the village. The second option would be to use the seed-cake to produce biogas in a village-size digester and to use the biogas in a village biogas engine, e.g. to generate electricity, while selling the oil as by-product to the biodiesel factory. Thus, the village based scenario defines ten service pathways: plant oil stove V1 and V2, biogas lamp V1 and V2, biodiesel motor V1 and V2, plant oil engine V1 and V2, and biogas engine V1 and V2. In the regional scale Jatropha scenario (shown in Fig. 6) the depulping of the Jatropha fruit is carried out like in the village scenario, but the seeds are directly transported to the biodiesel factory 382 M. Gaul / Applied Energy 101 (2013) 376–392 Fig. 5. Jatropha scenario – village scale. Fig. 6. Jatropha scenario – regional scale. for further processing. The factory can use a larger press and a larger and more efficient diesel generator, but also has a tripled transport volume of seeds instead of oil. In addition, the factory can use a biogas digester.9 Thus, the region-based scenario defines the two service pathways: biodiesel engine R and biogas engine R. Figs. 7–13 the results both for the ‘present case’ and a ‘best case’ are summarised. While the two sections on energy and cost analyses focus on the present case, the best case is discussed in the following section on sensitivity and uncertainty analysis. 3.1. Energy efficiency 3. Results The five different scenarios describe three baseline pathways and four competitive RET pathways, as well as 22 pathway configurations for the Jatropha scenarios as shown in Table 2. The results of the energy and cost analyses are presented below in detail. In 9 However, to fully make advantage of the produced biogas, the generation and feed-in of electricity would be the most profitable option that would require a feed-in agreement with the power utility PLN. In Fig. 7, the CED is displayed in kWh/kWh (kWh/kLmh in the case of lighting). The four baseline scenarios have a CED between 5.9 kW h for the 3-stone fire and 8.8 kW h for the diesel engine per supplied kWh useful energy and 11.7 kW h for the kerosene lamp per supplied kLm h light. While the CED of the 3-stone-fire can be considered as renewable (as long as the wood resources are sustainably used) the kerosene and diesel CED are almost 100% fossil. Comparing the energy service pathways of the baseline with the three Jatropha scenarios the steep increase of CED is the 383 M. Gaul / Applied Energy 101 (2013) 376–392 Table 2 The 29 analysed energy service pathways. Scenario Cooking (9 pathways) Lighting (8 pathways) Mechanical power (12 pathways) Baseline (3 pathways) 3-stone fire Kerosene lamp Diesel motor Jatropha (22 pathways) Plant oil stove H1/H2/V1/V2 Biogas stove H1/H2 Plant oil lamp H1/H2/V1/V2 Biogas lamp H1/H2 Plant oil engine V1/V2 Biogas engine V1/V2/R Biodiesel engine H1/H2/V1/V2/R Alternative RET (4 pathways) Solar stove Improved wood stove LED solar lamp Hydro turbine Table 3 Variation of selected input parameters and definition of the best case. a b c d Input parameter Unit Present case Variation Best case Variation 1: Capital costs Average interest rate % 16.8 8.4–33.4 8.4 Variation Kerosene Kerosene Kerosene IDR/l IDR/l IDR/l 2500 3500 5500 9000 9500 13,000 9000 9500 13,000 Diesel – retail Diesel – road Diesel – village IDR/l IDR/l IDR/l 4500 5000 8000 9000 9500 13,000 9000 9500 13,000 Variation 3: Manual processing Harvesting of fruit Cleaning of seeds Pressing of oil kg/wd kg/wd l/wd 30 8 2 60 16 4 60 16 4 Variation 4: End-use efficiency Efficiency of a plant oil stovea Efficiency of a plant oil lamp % Lm/W 39 0.1 50 No 50 0.1b Efficiency of a biogas stove Efficiency of a biogas lamp Average efficiency of small engines < 5 kW % Lm/W % 55.6 0.75 13.35 No No 26.7 55.6c 0.75c 26.7d Variation 5: Transport cost Average distance biodiesel factory – Jatropha farmer Average distance biodiesel factory – fuel depot km km 150 150 75 75 75 75 2: Fuel costs – retail – town – village Comparison of a wick stove with a pressurised plant oil stove stove. No high-efficiency lamps available. No further increase possible. Improved engine quality and operation. Fig. 7. Comparison of the cumulated energy demand. 384 M. Gaul / Applied Energy 101 (2013) 376–392 Fig. 8. Comparison of the cumulated fossil energy demand. Fig. 9. Comparison of the net energy balance. most obvious difference. The total CED of the Jatropha oil service pathways for cooking is 40–60% higher compared to the 3-stone fire. This ratio increases to 170–280% in the case of mechanical power, while in the category of lighting the CED increases by 170–310%. If only the fossil share of the CED is considered (see Fig. 8), the picture changes greatly. The increased fuel demand for machines and transport accumulates to significant fossil CED values, especially for biodiesel production and the V2 and R pathways. With regards to centralised biodiesel production (Biodiesel motor R) the fossil CED reaches almost 5 kW h, which is 56% of the diesel engine baseline value. The household Jatropha service pathways that apply only manual processing have the lowest fossil energy demand. However, with exception of the solar stove, all competitive RET pathways show even lower values than the best performing Jatropha pathways for both renewable and fossil CED. In Fig. 9, the NEB is displayed in kWh (kLmh in the case of lighting). Analysing the NEB values, we can see that there is no linear relation between the (fossil) CED and the NEB, as the NEB describes mainly the extent of required auxiliary energy. The kerosene and diesel-based baseline pathways still have a NEB of about 0.85, while the 3-stone fire has a NEB of 1 as no auxiliary energy is involved in the manual wood–fuel chain. We can see that the plant oil-based service pathways can reach a NEB even higher than 0.85 if the processing is mainly manual as in the household scenario and the V1 pathways, while the V2 pathways with enginepowered hulling and all biodiesel pathways have a reduced NEB. Biogas has almost the same NEB as comparable plant oil pathways because the higher cumulated auxiliary energy demand of the concrete digester is compensated due to the allocation of plant oil as a by-product. The net energy ratios of the 29 pathway configurations are displayed in Fig. 10 in kW h/kW h (kLm h/kW h in the case of lighting). The gross energy ratio has not been displayed as it shows a similar characteristic with only slightly higher values. The baseline pathways have a NER of 0.17, 0.08 and 0.11 for cooking, lighting and mechanical power respectively. All Jatropha oil service pathways have an even lower NER compared to these baseline values. The biogas pathways have a higher NER due to the allocation of M. Gaul / Applied Energy 101 (2013) 376–392 385 Fig. 10. Comparison of the net energy ratio. Fig. 11. Comparison of labour demand. plant oil as a by-product and, in the case of lighting, caused by the 7.5 times higher luminous efficiency of biogas lamps compared to kerosene wick lamps. Since a highly efficient LED has a 133 or 1000 times higher luminous efficiency than a biogas lamp or kerosene wick lamp respectively, the NER of the solar LED lamp is therefore as high as 8.6 kLm h/kW h. While the energy input in terms of human labour was excluded from the energy analysis, the labour intensity of the service pathways is presented in Fig. 11 to complement the results of the energy analysis. The high NEB of the household pathways is paid for with a high labour intensity that reaches prohibitive values in the case of plant oil for lighting and biodiesel for mechanical power. Without entering the debate on human metabolic rates and efficiencies, it can be stated that it is hardly a reasonable value adding activity to provide 1 kW h of mechanical power with a physical labour input of 8 h via a long and complex conversion pathway. 3.2. Cost efficiency In Fig. 12 the net energy costs (NEC) are presented in € per kW h useful energy output (€ per kLm h in the case of lighting). For the household scenarios, the net energy costs of Jatropha oil and biodiesel service pathways exceed the costs of the baseline by the factor 4–15. For the village scenarios it is still 30–250% for cooking and lighting, while in the case of mechanical power from plant oil the costs are only 90% of the baseline value. The costs of biogas services exceed the baseline in the case of cooking by the factor 6– 9, while for lighting and mechanical power the costs are only 47% and to 73% of the respective baseline values. In the household scenario, biogas for lighting and mechanical power, as well as plant oil for mechanical power are cost competitive, while in the village and regional scenarios also plant oil for lighting as well as biodiesel might become cost competitive in the nearby future. However, 386 M. Gaul / Applied Energy 101 (2013) 376–392 Fig. 12. Comparison of the net energy costs. Fig. 13. Comparison of the return on labour. the competitive RET service scenarios all have lower specific energy costs per kW h or kLm h. All values of the NEC presented include labour costs of €1.95 per workday.10 In Fig. 13 the return on labour is shown in € per workday for all service pathways. The ROL for lighting and mechanical power are calculated based on the avoided energy costs of the kerosene lamp and diesel engine of the baseline scenario. Because the 3-stone fire has no costs besides labour, for cooking the avoided energy costs of a kerosene stove of €0.12 per kWh is used instead. As the kerosene stove is used only by a small fraction of the rural population, the results for the return on labour of cooking stoves have to be interpreted with care. Of the 29 service pathways, only the improved stove, the biogas lamps and the plant oil and biogas engines reach the benchmark of €1.95. However, biogas lamps might not be relevant for illumination levels below 1480 kLm h per year as provided by the biogas lamps and the ROL for improved stoves cannot be 10 For qualified labour (e.g. at the biodiesel factory) the salary is doubled to €3.90 per workday. achieved by poor households that do not already use kerosene for cooking. 3.3. Sensitivity and uncertainty analysis A life cycle assessment includes a large quantity of input data, involving many estimates and assumptions (see Tables 4–7). Obviously, a sensitivity and uncertainty analysis cannot be conducted for more than 100 input parameters and 29 RESPs. Significant parameters were therefore identified according to the following two criteria: there exists some debate about the correct parameter value and/or a strong influence of the parameter on the results is expected. Five groups of input parameters have been selected:  capital costs (average interest rate corrected by inflation), with main effects on the more complex RESPs that involve mechanical processing; M. Gaul / Applied Energy 101 (2013) 376–392 387 Table 4 Key input parameters for all five scenarios. Parameter Data Unit Source/remarks Inflation rate (July 2009 to July 2010) Average interest rate (corrected by inflation) Average exchange rate February 2010 Average non qualified/qualified labour cost on Sumbawa Operation time for stoves Operation time of kerosene and plant oil lamps Operation time of biogas and solar LED lamps Operation time of motors (mechanical power as end-use) Operation time of machines, motors and refinery Lifetime of stoves, lamps, motors and machines Ship transport fossil fuels Tank truck transport fuels Bus/pick-up transport fuels Light transport Jatropha Pick-up transport Jatropha Tank truck transport Jatropha biodiesel CED steel CED concrete CED clay CED aluminium CED glass CED plastic CED silicon LHV Crude oil LHV Diesel LHV Kerosene LHV Petrol LHV Firewood LHV Jatropha fruits LHV Jatropha seeds LHV Jatropha oil LHV Jatropha press cake LHV Jatropha biogas LHV Jatropha biodiesel 6.22 16.8 12819 1.95/3.90 1000 2190 1850 1000 2000 5 1200 150 20 10 150 150 5.5 0.25 1.27 52.14 3.3 20 127.74 11.11 11.85 12.14 11.99 4.06 5.89 7.08 11.31 5.7–6.0 5 10.36 % % IDR/€ €/wd h/a h/a h/a h/a h/a years km km km km km km kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg kW h/kg BPS [50] Bank of Indonesia [56], average of national and rural rates AGB [57] Based on 10 villages on Sumbawa, 1 workday (wd) is assumed as 8 h Estimate, 2.74 h per day Estimate, 1.5 lamps for 4 h per day Estimate, 5 h per day Estimate, 2.74 h per day Estimate, 8 h per day, 4.8 days per week Estimate, clay stoves 1 year and improved stove 3 years lifetime Distance of refinery in Surabaya/Java region to fuel depots on Sumbawa 2 fuel depots, 15 filling stations and 8 kerosene distributors Private transport to village with difficult road conditions farmer field G village coordinator coordinator G biodiesel factory Biodiesel factory – Pertamina depot Based on GEMIS data for steel mix Based on GEMIS data for concrete Based on GEMIS data for clay roofing tile Estimate, based on GEMIS data for aluminium Estimate, based on GEMIS data for window glass Estimate, based on average GEMIS data for different plastics Estimate, based on average GEMIS data for silicon Based on GEMIS data Based on GEMIS data Based on GEMIS data Based on GEMIS data Air dry wood (moisture content 20% wet basis) [58] Openshaw [59] Openshaw [59] Openshaw [59] Calculated, depending on oil extraction ratio Syaharuddin et al. [60] Prueksakorn et al. [61] Table 5 Key input parameters for the baseline scenario. Parameter Data Unit Source/remarksxxx Average daily time of firewood collection Average daily time of wood chopping and drying Efficiency of 3-stone fire Share of imported crude oil (2008) Price of kerosene (village/town/retail) Price of diesel (village/road/retail) Cooking efficiency of a wick kerosene stove Power of a 3-stone-fire Power of a wick kerosene stove Lighting efficiency of a wick kerosene lamp Power of a wick kerosene lamp Average efficiency of a small engine < 5 kW Cost of a diesel generator 1.15 0.35 17 27.4 5500/3500/2500 8000/5000/4500 38.2 3.0 1.2 0.1 40 13.35 109 h/day h/day % % IDR/l IDR/l % kW kW Lm/W Lm % €/kW Based on 10 villages on Sumbawa/Indonesia Based on 10 villages on Sumbawa/Indonesia Estimate based on MacCarty et al. and the WBT 3.0/4.1 [62–64] [65] Price of February 2010 (village: €0.43/l) based on local survey Price of February 2010 (village: €0.62/l) based on local survey Calculated with data from MacCarty et al. [62] Estimate, based on MacCarty et al. [62] Estimate, based on MacCarty et al. [62] Van der Plas [66] Van der Plas [66] Estimate, based on local data on fuel consumption Local price of February 2010  fossil fuel costs, with main effects on all energy and transport intensive RESPs;  efficiency of manual harvest and processing, with main effects on the labour intensive RESPs;  efficiency of end-use appliances, with main impacts on RESPs for mechanical power; and  transport requirements (average distance of biodiesel factory to the Jatropha farmers and to the fuel depot) with main impacts on the biodiesel and the regional biogas service pathways. The auxiliary energy demand of materials was not further investigated as its proportion of the total CED remains for all ser- vice pathways below 1%.11 These five variations were first analysed one at a time and then combined to illustrate the possible ‘best case’ for all scenarios as defined in Table 3 and demonstrated by the dark bars in the Figs. 7–13. Because very conservative data and estimations were used for the present case, present and best case span at the same time the range for the main effects of data uncertainties. A variation of the interest rate between 8.4% and 33.6% shows the main impact on the capital intensive pathways such as the so11 Exceptions are the solar LED lamp (12.2%) the solar stove (15.2%) and the microhydropower plant (5.3%), which are only used as a reference for comparison and even the doubling of their energy demand for materials would not influence the results of the analysis. 388 M. Gaul / Applied Energy 101 (2013) 376–392 Table 6 Key input parameters for the competitive RET scenario. Parameter Data Unit Source/remarksxxx Cost of MHP investment Annual operation time of MHP Lifetime of MHP concrete structure/ turbine Lighting efficiency of a solar led lamp Cost of a solar lamp Cost of a solar stove Power of a solar stove Cost of an improved stove Cooking efficiency of an improved stove 3000 5000 20/10 100 60 50 0.15 30,000 26.8 €/kW h/a years Lm/W € € kW IDR % MHPP Indonesia [67] Estimate, depending on maintenance and water availability MHPP Indonesia [67] Based on the Solux 105 [54] €40 for the Solux 105 lamp + 50% for tax and retail Estimate for Sumbawa, based on GEMIS 4.6 data Meyer [52] Estimate for Sumbawa Estimate, based on MacCarty et al. [62] Table 7 Key input parameters for the bioenergy scenarios. Parameter Data Unit Source/remarks Harvesting fruit Cleaning seeds (separating seeds and fruit pulp) Cleaning seeds with mechanical huller Cost of mechanical huller Ratio of clean seeds to harvested fruit Extraction rate of a manual Jatropha press Production rate of a Jatropha ram press Cost of a manual Jatropha ram press Biogas production per kg Jatropha press cake Density of Jatropha Biogas Biodiesel production rate (biodiesel per plant oil input) Biodiesel production capacity Cost of a wick plant oil stove Cooking efficiency of a wick plant oil stove Power of a wick plant oil stove Cost of a pressurised plant oil stove Cooking efficiency of a pressurised plant oil stove Power of a pressurised plant oil stove Cost of a biogas stove Cooking efficiency of a biogas stove Power of a biogas stove Cost of a household size biogas digester Cost of a wick plant oil lamp Lighting efficiency of a wick plant oil lamp Cost of a biogas lamp Lighting efficiency of a biogas lamp Power of a biogas lamp Cleaning seeds with a motor huller Cost of a motor huller Extraction rate of a small motor screw press Production rate of a small motor screw press Cost of a small motor screw press Cost of a medium size biogas digester (70 kW gas) Extraction rate of a large motor screw press Production rate of a large motor screw press Cost of a large motor screw press Selling price for plant oil at farm gate 30 8 50 500,000 70 0.22 2 5 Mio 0.5 1.2 0.918 1275 150,000 38.2 1.2 44 50 2 300,000 55.3 1.2 8 Mio 50,000 0.1 175,000 0.75 800 533 15 Mio 0.27 200 50 Mio 500 0.27 320 80 Mio 5000 kg/wd kg/wd kg/wd IDR % l/kg l/d IDR m3/kg kg/m3 MJ/MJ l per day IDR % kW € % kW IDR % kW IDR IDR Lm/W IDR Lm/W Lm kg/wd IDR l/kg l/wd IDR G/kW l/kg l/wd IDR IDR Average of 10 villages Average of 10 villages Data of February 2010, Sumbawa island Data of February 2010, Sumbawa island [59] Data of February 2010, Sumbawa island Considering 3 times pressing of seeds for optimum extraction rate Data of February 2010, Sumbawa island Syaharuddin et al. [60] Estimate by generic biogas density [68] 1.06 l biodiesel per l plant oil (20% methanol input) Data of February 2010, Sumbawa island Estimate, design as of a kerosene wick stove Estimate, equal to a kerosene wick stove Estimate, equal to a kerosene wick stove Estimate, based on BSH estimation of a local production cost of 50 US$ [52] 45–55%, based on the BSH stove [55] 2–2.5 kW, based on the BSH stove [55] Data of February 2010, Sumbawa island Estimate, equal to LPG stove efficiency [62] based on the LHV of biogas Estimate, based on MacCarty et al. [62] Estimate for Sumbawa based on local costs and [69] Data of February 2010, Sumbawa island Estimate, based on kerosene wick lamp efficiency Based on cost of a pressurised kerosene lamp Adopted from an Indian biogas lamp [70] Adopted from an Indian biogas lamp [70] Data of February 2010, Sumbawa island Estimate Data of February 2010, Sumbawa island Data of February 2010, Sumbawa island Data of February 2010, Sumbawa island Estimate based on GEMIS 4.6 values Data of February 2010, Sumbawa island Estimate Estimate Estimate, allocated as positive income to the biogas pathways lar stove, solar LED lamp and the hydropower turbine that have no or low labour content. Reducing the interest rate by 50% translates into a decrease of the net energy costs of up to 34%, while doubling the interest rate increases the net energy costs by up to 77%. However, the impact on Jatropha oil-based pathways is below ±5%. Also the change for both the capital and labour-intensive pathways for biodiesel still remains between 6% and +2%, while the more capital-intensive biogas for mechanical power pathways reach 16% and +38%. The maximum resulting change in the return on labour is between €0.27 and €0.62 per workday, with the largest changes occurring for the already competitive pathways, which maintain a ROL above the benchmark of €1.95 per workday even with doubled capital costs, while the other pathways do not reach the benchmark also with reduced capital costs. In Indonesia, fossil fuels are highly subsidised. If the official kerosene and diesel prices would be increased by about 250% and 100% respectively, the resulting rural kerosene and diesel prices would reach about IDR 13,000 (about €1 per litre). As a consequence, the NEC for kerosene cooking and lighting, as well as diesel-based mechanical power, would increase almost proportionally. The net energy costs of other pathways increase up to 25% depending on the share of fossil fuels they require for transport and machines. Compared to a kerosene cooker, the ROL increases significantly, making plant oil cooking in the village scenario cost competitive with kerosene cooking. In the case of lighting the ROL of biogas service pathways further increases to €6.24 and €9.75 per workday and the two plant oil-based pathways V1 and V2 reach a competitive ROL with about €2.50 per workday. For mechanical power, now the biodiesel pathways (V1, V2 and M. Gaul / Applied Energy 101 (2013) 376–392 RE) also reach a ROL of about €2.10 to €2.20 per workday, while the plant oil and biogas-based service pathways increase their return on labour to €3.50 and €4.30 per workday. The locally observed harvest performance of 30 kg of fruit per workday might be low compared to more intensive cropping systems. Even more critical are the observed low efficiencies for manual hulling (8 kg seeds per workday) and pressing (2 l plant oil per work day) of seeds. If the manual harvest and processing performance could be doubled, the net energy costs for all Jatropha pathways could be reduced to between 19% and 56%. The largest impact can be observed for the village and regional pathways, where plant oil and biogas-based mechanical power would then reach a ROL of €3.20 to €5.30 per workday. By contrast for the household pathways the doubling of manual harvest and processing performance indeed reduces the net energy costs significantly, but the difference is still too low to make the pathways competitive with a ROL below €0.94 per workday for plant oil cooking and lighting or biodiesel and below €1.79 per workday for biogas cooking. While the first three variations only impacted the costs, an improvement of the end-use appliances has a direct influence on both energy and cost efficiency. For cooking, the use of a pressurised plant oil cooker (e.g. the Protos cooker12) with an efficiency of 50% is assumed. For lighting an increase of lighting efficiency for plant oil lamps is not feasible as a pressurised plant oil lamp (comparable to a pressurised kerosene lamp) is currently not available. For mechanical power, the use of biodiesel, plant oil and biogas engines with a doubled efficiency of 26.7% are assumed at a 50% higher price and compared with an equally improved diesel engine baseline. For cooking the CED and NEC can be reduced by about 23% and the ROL increased by up to 52%. However, the NEC remains three times higher as the baseline in the best case and the ROL is – at €1.73 per workday – still below the benchmark. Doubling the engine efficiency also halves the CED and the NEC, however, because the baseline is improved proportionally, the difference of the NEC to the baseline and the ROL remains unchanged. The CED of transport is surprisingly homogeneous for all transported energy carriers, with less than 0.01 kW h per transported kW h fuel.13 Because transport represents less then 2.5% of the cumulated energy demand of the Jatropha pathways, the impact of a reduced transport distance is limited. If the distance between the biodiesel factory and farmers and the fuel depot would be halved to 75 km each, the total CED of the biodiesel pathways would only decrease in the range of 0.4–1.7%. If only the fossil CED is considered, the decrease is in the range of 4–18% compared to the present case. The reduced energy demand for transport increases both the NEB and NER by 4–14%, with the largest effects for the biodiesel and biogas engines in the regional scenario. The impact on net energy costs remains mainly in a reduction between 0.5% and 3.6%, but reaches 13% for the biogas engine in the regional scenario. This can be easily explained as for a biogas production at regional scale not only the plant oil but the whole seed need to be transported to the biogas plant, which triples the transport volume. In such a scenario, a reduction of the transport distance causes the highest impact. The ROL also increases by 7–17%, but both NEC and ROL do not change sufficiently to change the analysis made above. While the variations of isolated parameters mostly make no sufficient impact to significantly change the results of the present case, a combination of the five parameters can cause larger effects. The combination of the five variations as ‘best case’ is indicated by the slim dark bars in Figs. 7–13. Regarding the energy analysis, the general picture has not significantly changed. The strong reduction of cumulated energy demand for mechanical power pathways oc12 BSH Bosch und Siemens Haushaltsgeräte GmbH is introducing its Protos pressurised plant oil cooker to Indonesia [55]. 13 This value already includes the different transport distances and vessels. 389 curs equally for the baseline without increasing the competitiveness of the RESPs. The slightly increased net energy balances for the biodiesel and the regional biogas pathway demonstrates the impact of the halved transport distance of 75 km between the biodiesel factory and the farmers and biodiesel depot. The labour costs – as the largest share of the net energy costs – also strongly decreased, resulting in a significant increase of ROL. Because the fuel costs for the fossil baseline also increased, the strongest impact can be observed in the category of lighting, where now even the plant oil lamp (V1 and V2) can compete with the kerosene baseline. However, the solar LED lamp remains by far the most attractive alternative for lighting. In the category of cooking, in particular the village plant oil pathways V1 and V2 become competitive compared to kerosene stoves, but not to wood cooking. Both the solar and improved stoves remain the most attractive options for cooking. In the category of mechanical power all village and regional Jatropha pathways become competitive compared to the baseline both in NEC and ROL. However, hydropower is the more attractive option if available and the high CED remains a problem, especially for the biodiesel pathways with a NEB below 0.5. 4. Discussion In the category of cooking the energy analysis reveals a low performance of Jatropha-based energy service pathways compared to the baseline or the competitive RET pathways. This picture changes if only fossil energy is considered. But the low NER causes high costs and these high costs make Jatropha-based cooking unattractive compared to a simple 3-stone fire and even more so compared to a solar or an improved stove as long as firewood can be easily collected. However, in the case of firewood scarcity, improved wood fuel stoves could use wood from fast growing scrubs that could be planted as living fences instead of Jatropha, thus equally reducing the local pressure on wood resources, but requiring less effort compared to the Jatropha harvest and processing. For lighting, Jatropha oil lamps can only compete with the very inefficient kerosene wick lamp. The problem of biogas lamps is that even though they provide about 20 times as much light as kerosene lamps, they have only an about seven times higher lighting efficiency, thus increasing the net annual lighting costs by the factor of three. This effect is similar for pressurised kerosene lamps which are therefore rarely used on Sumbawa. Additionally, a biogas digester and lamp require a very high initial investment and it can be expected that biogas lamps will only be affordable to richer households, which will probably choose a solar home system instead. A small solar LED lamp, on the other hand, still provides five times as much light as a wick kerosene lamp but with an up to a 1000 times higher lighting efficiency. Therefore a solar LED lamp shows a far better characteristic and potential as a lighting source, even if the introduction and distribution of such lamps in rural areas surely poses a challenge. For the supply of mechanical power, Jatropha-based pathways show potential but the specific pathways need to be further analysed, before any specific Jatropha-based strategy can be supported. Such a comprehensive analysis would need to look into the four sub-systems (extraction, conversion, distribution and end-use) of the service pathways to discuss possible scope for optimisation. Therefore, two critical issues need to be addressed in further investigations: the problem of long supply chains with low conversion efficiencies and the balancing of capital and auxiliary energy input versus labour intensity. Comparing the energy service chains of the fossil baseline with Jatropha-based pathways reveals a critical difference: while the fossil supply chain is the longer and more complex one, it has during the extraction, conversion, and distribution phase very high en- 390 M. Gaul / Applied Energy 101 (2013) 376–392 ergy ratios of above 90%. This is caused by the economy of scale for the established fossil supply chain, while the greatest losses occur only at the end-use level, multiplying the CED of the final energy by a factor 3–10. The Jatropha chain might be shorter and simpler, but the NER during extraction and conversion are only between 20% and 50% and similarly low as the fossil pathways at the enduse level. Consequently, the overall NER decreases to a few per cent, as all primary and auxiliary energy inputs are divided several times by these low efficiencies. Therefore the Jatropha-based service pathways show a high CED including significant amounts of fossil fuels as auxiliary energy. If these auxiliary fossil fuels would also be replaced by Jatropha-based fuels, the CED would multiply again, causing a strong increase of resource demand and costs and reducing the already low competitiveness. Of all Jatropha-based service pathways, the household-based scenarios H1 and H2 show the highest NEB because of the low mechanisation of the processing. However, this advantage considering the energy and capital intensity is paid for with a high labour intensity due to the manual picking, hulling and pressing of seeds that cause a very low ROL. In fact, the interviewed farmers criticised the low ROL of picking and manual cleaning of Jatropha seeds, and the acceptance of manual pressing seems more than doubtful. On the other hand, a higher mechanisation reduces the NEB and can increase the NEC if the reduced labour costs do not outweigh the increased capital and fuel costs. The introduction of a cheap manual hulling tool decreases the cost without any impact on the NEB (comparison of H1 and H2), while the replacement of the manual tool with a engine-powered huller cannot further decrease the NEC while the NEB is decreased significantly due to the required fuel for the engine (comparison of V1 and V2). While a reduction of drudgery can be a justification besides cost reduction, the impact of the increased auxiliary energy demand needs to be considered. Due to the mechanisms described above, the energy demand for transport can accumulate significant values reaching up to 68% of the total auxiliary energy demand in the case of the regional biogas scenario. The highest amount of auxiliary energy is required for the regional biodiesel engine pathway in the present case: in order to provide 1 kW h mechanical power, 5 kW h auxiliary fossil energy are required of which 54% are methanol, 17% are the energy demand for transport, 1% the energy demand of materials, while the remaining 28% are the engine fuels for hulling, pressing and power for the biodiesel factory. The required auxiliary energy can be halved to 2.5 kW h in the best case, however, the total CED remains also in the best case substantial, with almost 15 kW h. Considering these points, the results indicate the best potential for the village-based Jatropha scenario, which is using both the Jatropha oil and the biogas from the Jatropha press-cake to locally provide mechanical power or also electricity. The export of Jatropha oil as proposed in the scenario would make only sense for surplus production after all local demand is satisfied. pete with wood-based cooking as long as the supply of firewood can be collected (or planted) locally. In the case of lighting, some Jatropha service pathways can compete with kerosene lighting, but current developments in LED lighting show a far higher potential in terms of energy and cost efficiency. Only Jatropha-based mechanical power shows a potential in the energy and cost analyses, especially in areas with insufficient hydropower potential. Of these, non-mechanised Jatropha pathways at household-scale fail the economic analysis because of a prohibitive low return on labour. Even so, Jatropha, and bioenergy in general, is often promoted for its potential to increase rural employment, such an employment must also create a sufficient and competitive return for farmers. A promotion of householdscale Jatropha pathways therefore risks to produce negative development effects by advising smallholder farmers to invest in unproductive side-crops. Also the biodiesel-oriented pathways show a low energy and cost performance compared to plant oil and biogas pathways. This is mainly caused by the high auxiliary energy input for methanol, engine fuels and transport as well as the fuel and capital costs. While the performance of the small-scale biodiesel production system could be further improved, its viability considering the supply of rural areas remains questionable: the operation and maintenance of a biodiesel factory in an rural environment is doubtless challenging, but (as shown by the present entrepreneur on Sumbawa) it could be done at a (to a limited extend) regional scale transporting the plant oil (or the seeds) from and the biodiesel (and the press cake) to the surrounding villages. At such a scale, the operator can develop the skills and capacity required to maintain the biodiesel factory, but then would be more interested in securing stable long-term contracts to sell the biodiesel with lower transaction costs to the national market, while transporting more profitable goods to the villages then press cake. Village-scale Jatropha pathways on the other hand could be a more energy and cost efficient strategy. This is especially true, if the energy and cost performance of the system is further increased due to the combined production and use of plant oil and biogas with a dual fuel engine to locally provide mechanical power and electricity. But even if the operation of a village mini-grid powered by such a dual fuel engine might be technically less demanding then a biodiesel factory, the technical and logistical operation is still challenging especially if fully operated at village scale. In sum, policy makers are advised to consider the pros and cons of Jatropha-based rural energy service pathways in their individual context before supporting Jatropha in their local constituencies. More research is required to study the feasibility of village-scale Jatropha pathways compared to other small-scale bioenergy service pathways. And with the current developments and cost reductions of photovoltaic systems and a finally developing small wind power market, also these technologies should be included into future analyses. Acknowledgements 5. Conclusion This research highlights the importance of taking the farmers’ labour and profit into any analysis of agriculture-based bioenergy systems as well as to compare the possible uses of a specific rural bioenergy service pathway with the existing and other possible alternatives before investing into a specific crop. The presented results paint a sobering picture of the technical and financial feasibility of small-scale Jatropha-based energy services for rural areas. A low energy efficiency of the Jatropha processing chain together with high labour intensities result in high costs and a low return on labour for many of the studied service pathways. Jatropha-based service pathways therefore cannot com- The research has been kindly supported by the project owner Ibrahim Syaharuddin and Dr. Tatang Hernas Soerawidjaja from the Institute of Technology Bandung and was funded by the Hans Böckler Foundation. References [1] IEA. World Energy Outlook 2002. Paris: International Energy Agency (IEA); 2002. ISBN 978-9264198357. [2] IEA. World Energy Outlook 2010. Paris: International Energy Agency (IEA); 2010. ISBN 978-9264086241. [3] AEGCC. Energy for a sustainable future: summary report and recommendations. Vienna: Secretary-General’s Advisory Group on Energy M. Gaul / Applied Energy 101 (2013) 376–392 [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] and Climate Change, United Nations; 2010. [accessed 09.12.10]. WBGU. Welt im Wandel: Zukunftsfähige Bioenergie und nachhaltige Landnutzung. Berlin: Wissenschaftlicher Beirat der Bundesregierung Globale Umweltveränderungen; 2009. ISBN 978-3936191219. IPCC. Renewable energy sources and climate change mitigation. Special report of the intergovernmental panel on climate change. Cambridge (UK): Cambridge University Press; 2011. ISBN 978-1107607101. Legros G, Havet I, Bruce N, Bonjour S. The energy access situation in developing countries. New York (NY): United Nations Development Programme; 2009. [accessed 27.11.09]. Reiche K, Covarrubias A, Martinot E. Expanding electricity access to remote areas: off-grid rural electrification in developing countries. In: WorldPower 2000. London: Isherwood Production; 2000. p. 52–60. [accessed 17.11.09]. REN21. Renewables 2011: global status report. Paris: REN21 Secretariat; 2011. [accessed 10.08.11]. WEC/FAO. The challenge of rural energy poverty in developing countries. London: World Energy Council (WEC) and Food and Agriculture Organization of the United Nations (FAO); 1999. [accessed 20.05.06]. Meyer R, Börner J. Bioenergieträger – eine Chance fur die Dritte Welt: Verfahren – Realisierung – Wirkungen. Berlin: Edition Sigma; 2002. ISBN 9783894048204. Karekezi S, Kusum L, Coelho ST. Traditional biomass energy – improving its use and moving to modern energy use. Bonn (Germany): Secretariat of the International Conference for Renewable Energies; 2004. [accessed 25.07.06]. Sagar AD, Kartha S. Bioenergy and sustainable development? Ann Rev Environ Resour 2007;32(1):131–67. http://dx.doi.org/10.1146/annurev.energy.32. 062706.132042. Henning RK. The Jatropha system economy & dissemination strategy. In: International conference for renewable energies, Bonn Germany; 2004. WorldCat. Results for ‘Jatropha curcas’ in the category ‘article’; 2012. [accessed 06.01.12]. Heller J. Physic nut – Jatropha curcas L.. Rome: International Plant Genetic Resources Institute; 1996. ISBN 978-9290432784. Gübitz G, Mittelbach M, Trabi M. Biofuels and industrial products from Jatropha curcas. Graz: Dbv Verlag fur die Technische Universität Graz; 1997. ISBN 978-3704102423. Jongschaap R. Claims and facts on Jatropha curcas L. – Global Jatropha curcas evaluation, breeding and propagation programme. Wageningen: Plant Research International; 2007. [accessed 27.02.10]. GEXSI. Global market study on Jatropha. London: GEXSI LLP; 2008. [accessed 27.12.10] Jongh J. The Jatropha handbook: from cultivation to application. 1st ed. Eindhoven: FACT Foundation; 2010. ISBN 978-9081521918. Brittaine R, Lutaladio N. Jatropha: a smallholder bioenergy crop – the potential for pro-poor development. Rome: Food and Agriculture Organization of the United Nations; 2010. ISBN 978-9251064382. NL Agency. Jatropha assessment – agronomy, socio-economic issues, and ecology. Utrecht: Netherlands Programme Sustainable Biomass; 2010. [accessed 02.10.11]. Sorda G, Banse M, Kemfert C. An overview of biofuel policies across the world. Energy Policy 2010;38(11):6977–88. http://dx.doi.org/10.1016/ j.enpol.2010.06.066. Timilsina GR, Shrestha A. How much hope should we have for biofuels? Energy 2010;36(4):2055–69. http://dx.doi.org/10.1016/j.energy.2010.08.023. Divakara B, Upadhyaya H, Wani S, Gowda CL. Biology and genetic improvement of Jatropha curcas L.: a review. Appl Energy 2010;87(3):732–42. http://dx.doi.org/10.1016/j.apenergy.2009.07.013. Mukherjee P, Varshney A, Johnson TS, Jha TB. Jatropha curcas: a review on biotechnological status and challenges. Plant Biotechnol Rep 2011;5(3):197–215. http://dx.doi.org/10.1007/s11816-011-0175-2. Koh MY, Mohd, Ghazi TI. A review of biodiesel production from Jatropha curcas L. oil. Renew Sustain Energy Rev 2011;15(5):2240–51. http://dx.doi.org/ 10.1016/j.rser.2011.02.013. Abdulla R, Chan ES, Ravindra P. Biodiesel production from Jatropha curcas: a critical review. Crit Rev Biotechnol 2011;31(1):53–64. http://dx.doi.org/ 10.3109/07388551.2010.487185. WEC. Comparison of energy systems using life cycle assessment. London: World Energy Council; 2004. ISBN 978-0946121168. Varun, Bhat I, Prakash R. LCA of renewable energy for electricity generation systems: a review. Renew Sustain Energy Rev 2009;13(5):1067–73. http:// dx.doi.org/10.1016/j.rser.2008.08.004. ISO 14040. Environmental management: life cycle assessment – principles and framework. ISO 14040, 2nd ed. Geneva: International Organization for Standardization; 2006. ISO 14044. Environmental management: life cycle assessment – requirements and guidelines. ISO 14044, 1st ed. Geneva: International Organization for Standardization; 2006. Benoist A, Dron D, Zoughaib A. A relevant LCA methodology adapted to biomass-based products. In: Empowerment of the rural actors a renewal of [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] 391 farming systems perspectives. Thiverval-Grignon; 2008. p. 659–66. ISBN 9782-7380-1252-4. Rowe R, Chapman J, Whittaker J, Howard D, Taylor G. A systematic review of life cycle assessments of bioenergy chains for heat, power and liquid transportation fuel. Comparat Biochem Physiol – Part A: Mol Integr Physiol 2008;150(3):S182–3. http://dx.doi.org/10.1016/j.cbpa.2008.04.486. Silva Lora EE, Escobar Palacio JC, Rocha MH, Grillo Ren ML, Venturini OJ, Almazn del Olmo O. Issues to consider, existing tools and constraints in biofuels sustainability assessments. Energy 2010;36(4):2097–110. http:// dx.doi.org/10.1016/j.energy.2010.06.012. Cherubini F, Strmman AH. Life cycle assessment of bioenergy systems: state of the art and future challenges. Bioresour Technol 2011;102(2):437–51. http:// dx.doi.org/10.1016/j.biortech.2010.08.010. GBEP. The global bioenergy partnership common methodological framework for GHG lifecycle analysis of bioenergy. Global Bioenergy Partnership, Food and Agriculture Organization; 2009. [accessed 08.06.11]. GBEP. GBEP sustainability indicators. Global Bioenergy Partnership, Food and Agriculture Organization; 2011. [accessed 08.06.11]. Farrell AE, Plevin RJ, Turner BT, Jones AD, OHare M, Kammen D. Ethanol can contribute to energy and environmental goals. Science 2006;311(5760):506–8. http://dx.doi.org/10.1126/science.1121416. Pleanjai S, Gheewala SH. Full chain energy analysis of biodiesel production from palm oil in thailand. Appl Energy 2009;86:S209–14. http://dx.doi.org/ 10.1016/j.apenergy.2009.05.013. Prueksakorn K, Gheewala SH. Full chain energy analysis of biodiesel from Jatropha curcas L. in Thailand. Environ Sci Technol 2008;42(9):3388–93. http:// dx.doi.org/10.1021/es7022237. Achten WM, Almeida J, Fobelets V, Bolle E, Mathijs E, Singh VP, et al. Life cycle assessment of jatropha biodiesel as transportation fuel in rural India. Appl Energy 2010;87(12):3652–60. http://dx.doi.org/10.1016/j.apenergy. 2010.07.003. Giampietro M, Mayumi K. The biofuel delusion: the fallacy of large-scale agrobiofuel production. London: Earthscan; 2009. ISBN 978-1844076819. Oeko Institut. Global emission model for integrated systems (GEMIS) version 4.6. Website; 2010. [accessed 09.12010]. UNEP. Towards sustainable production and use of resources: assessing biofuels. Paris: United Nations Environment Programme; 2009. ISBN 9789280730524. Reinhardt G. Screening life cycle assessment of jatropha biodiesel. Heidelberg (Germany): Institute for Energy and Environmental Research Heidelberg (IFEU) GmbH; 2007. [accessed 08.10.08]. Gmünder SM, Zah R, Bhatacharjee S, Classen M, Mukherjee P, Widmer R. Life cycle assessment of village electrification based on straight jatropha oil in Chhattisgarh, India. Biomass Bioenergy 2010;34(3):347–55. http://dx.doi.org/ 10.1016/j.biombioe.2009.11.006. Pearce D. Cost-benefit analysis and the environment: recent developments. Paris: Organisation for Economic Co-operation and Development; 2006. ISBN 978-9264010048. Wiskerke W. Towards a sustainable biomass energy supply for rural households in semi-arid Shinyanga, Tanzania. Master thesis. Utrecht (the Netherlands): Department of Science, Technology and Society, Utrecht University; 2008. [accessed 19.11.08]. BPS. Trends of the selected socio-economic indicators of Indonesia. 03230.0902. Jakarta: Badan Pusat Statistik Indonesia; 2009. [accessed 10.12.10]. BPS. BPS-Statistics Indonesia – strategic data 2010. 03220.1002. Jakarta: Badan Pusat Statistik Indonesia; 2010. [accessed 10.12.10]. Jungbluth N. Restricted life cycle assessment for the use of liquefied petroleum gas and kerosene as cooking fuels in India. Diploma thesis. Berlin: Technische Universitt Berlin; 1995. [accessed 12.12.10]. Meyer K. Vergleich verschiedener energieszenarios am beispiel einer ländlichen region in Indonesien. Diploma thesis. Aachen (Germany): RWTH Aachen; 2010. Braithwaite De. Fossil fuels at what cost? Government support for upstream oil and gas activities in Indonesia. Geneva: International Institute for Sustainable Development (IISD); 2010. [accessed 12.12.10]. Solux. Mobile solar lanterns. Website; 2010. [accessed 12.12.10]. BSH. Auf einen blick PROTOS. der pflanzenölkocher. BSH Bosch und Siemens Hausgeräte GmbH, München, Germany; 2008. [accessed 24.08.10]. Bank of Indonesia. The Indonesian banking statistics (SPI), vol. 8, no. 3, February 2010. Website; 2010. [accessed 10.11.10]. AGB. Association of German banks – currency converter. Website; 2010. [accessed 12.12.10]. 392 M. Gaul / Applied Energy 101 (2013) 376–392 [58] Kartha S, Leach G, Rajan SC. Advancing bioenergy for sustainable development – guideline for policymakers and investors, vols. I, II, and III. ESM 300/05. Washington (DC): Joint UNDP/World Bank Energy Sector Management Assistance Programme (ESMAP); 2005. [accessed 17.11.09]. [59] Openshaw K. A review of Jatropha curcas: an oil plant of unfulfilled promise. Biomass Bioenergy 2000;19(1):1–15. http://dx.doi.org/10.1016/S09619534(00)00019-2. [60] Syaharuddin I, Wiji A, Soerawidjaja T, Brodjonegoro T, Reksowardojo I. Produksi biogas dari bungkil jarak pagar (Jatropha curcas). Kelompok Studi Biodiesel, Institut Teknologi Bandung; 2009. [61] Prueksakorn K, Gheewala SH, Malakul P, Bonnet S. Energy analysis of jatropha plantation systems for biodiesel production in thailand. Energy Sustain Develop 2010;14(1):1–5. http://dx.doi.org/10.1016/j.esd.2009.12.002. [62] MacCarty N, Still D, Ogle D. Fuel use and emissions performance of fifty cooking stoves in the laboratory and related benchmarks of performance. Energy Sustain Develop 2010;14(3):161–71. http://dx.doi.org/10.1016/ j.esd.2010.06.002. [63] Bailis R, Oglel D, MacCarty N, Stil D. The water boiling test (WBT) – version 3.0. Berkeley: University of California, Berkeley; 2007. [accessed 26.09.10]. [64] ETHOS. The water boiling test (WBT) – version 4.1.2. Engineers in Technical and Humanitarian Opportunities of Service, Partnership for Clean Indoor Air; 2009. [accessed 25.07.11]. [65] CDI-EMR. Handbook of energy & economic statistics of Indonesia. Jakarta: Center for Data and Information on Energy and Mineral Resources of the Ministry of Energy and Mineral Resources (MEMR); 2010. [66] Van der Plas R. A comparison of lamps for domestic lighting in developing countries. Washington (DC): Household Energy Unit, World Bank; 1988. [accessed 17.11.09]. [67] Vetter S. Personal communication – mini hydro power project for capacity development, deutsche gesellschaft für technische zusammenarbeit (GTZ), Jakarta/Indonesia; 2010. [68] BINE. Biogas. 16; Bonn: Fachinformationszentrum Karlsruhe; 2003. [accessed 14.08.11]. [69] van Nes WJ. Feasibility of a national programme on domestic biogas in Indonesia. SNV Netherlands Development Organisation; 2009. [accessed 28.11.10]. [70] Mahapatra S, Chanakya H, Dasappa S. Evaluation of various energy devices for domestic lighting in India: technology, economics and CO2 emissions. Energy Sustain Develop 2009;13(4):271–9. http://dx.doi.org/10.1016/ j.esd.2009.10.005.