Ethiopia's emerging domestic biogas sector Current status, bottlenecks and drivers

Ethiopia experiences an energy and environmental crisis due to the sustained reliance on woody biomass to satisfy its energy needs. This situation could be improved by using biogas. This paper analyses the current status of the domestic biogas sector in Ethiopia and identi ﬁ es barriers and drivers that in ﬂ uence its development and further growth. The analytical framework used for the analysis combines the MultiLevel Perspective (MLP) and Strategic Niche Management (SNM). The information sources are: desk study through literature and internet research, online interviews and questionnaires with Ethiopian stakeholders, and an extensive ﬁ eld study including a large number of interviews with stakeholders inside and outside Ethiopia. The biogas sector in Ethiopia started with the launch of the National Domestic Biogas (NBPE) programme in 2008, which has led to the dissemination of over 8000 biodi- gesters so far, about 60% of what was initially intended. The use of domestic biogas has been triggered by the energy crisis in Ethiopia and the suitability of the technology with the physical geography. However, the dissemination has been affected by factors such as economic instability, poverty and illiteracy. Also, many Ethiopian farmers are trapped in a lock-in, where due to their limited purchasing power they cannot afford the niche technology; at least in the way it is being disseminated. Within the emerging biogas sector, the NBPE designated a diverse set of actors to contribute to the implementation of the niche technology. However, their alignment is poor and the private sector is not involved. Expectations have had to be lowered because targets were not met. Also, learning processes are not optimal. The paper is innovative because it provides an up-to-date review of status of and bottlenecks and drivers in the biogas sector in Ethiopia and it provides more insight


Introduction
The largest segment of the population in Ethiopia is located in rural areas and their main source of income comes from agriculture [1]. This same segment of the population represents the main energy consumer in Ethiopia and satisfies most of their energy needs with woody biomass [2,3]. The extensive demand for firewood has caused an energetic and environmental crisis since most of the forest coverage has been depleted over the last 35 years. This undermines firewood availability, soil fertility and the preservation of aquifers. Ultimately, Ethiopian farmers have to spend more resources (e.g. time or money) to have access to fuel wood. Meanwhile, their agricultural yield is reduced due to the lack of nutrients in the soil and the shortage of water [4]. Sundried and combusted livestock manure is sometimes used as an alternative energy source to fuel wood. However, this is not an improvement since the resulting 'dung cakes' have a low conversion efficiency to heat ( r 8%) [5] and inhibit the soil fertilisation that would have occurred if livestock manure would not have been combusted [6]. Domestic biogas, in contrast, also utilises organic waste such as livestock manure, but it can achieve a conversion efficiency to heat up to 55% and provides an enriched fertiliser called bio-slurry [5,7]. In 2008, the Ethiopian Government with the help of the SNV Development Organisation from The Netherlands launched the National Biogas Programme (NBPE) with the aim to up-scale the use of domestic biogas technology in the country. The NBPE started with a first implementation phase that concluded by the end of the year 2013 [4].
The main objective of this paper is to investigate the current status of domestic biogas (biogas used in households) in Ethiopia and to identify barriers and drivers that influence its development and further growth. The analytical framework applied in the paper, which will be further described in Section 2, combines the Multi-Level Perspective (MLP) and Strategic Niche Management (SNM).
The case study material presented in this paper was collected using three research methods: (1) desk study through literature and Internet research; (2) online interviews and questionnaires with relevant stakeholders; (3) an extensive field study including an extensive number of interviews with stakeholders inside and outside Ethiopia by the second author of this paper in the spring of 2014. Literature and Internet sources were used to collect qualitative information to analyse the relevant developments internal and external to the niche. The field study primarily consisted of open-ended, semi-structured interviews with 17 key informants: actors, experts and decision makers. The interviews focused on analysing the current stakeholder configuration within the sector, the status of the sector and the dynamics within it. Based on observations and a questionnaire, additional cultural and social factors that influence the sector were found. This paper is innovative in three respects. Firstly, it is the first paper that provides an up-to-date review of the case of the biogas sector in Ethiopia based upon primary data sources, mainly interviews. Although two recent publications, [8,9] do describe the status of the biogas sector in Ethiopia and Sub-Saharan Africa respectively, they base their findings on written data from 2011 and older, whereas the findings in the present paper are for the most part based on interviews from 2014. Also, these papers do not go into the drivers and barriers in the biogas sector in Ethiopia. In fact, one of the main recommendations for further research of [8] is to investigate empirically the factors that make the uptake of biogas technology in Ethiopia slow. Precisely that is what is aimed for in the present paper. Secondly, it is one of the first papers that apply a combination of the Strategic Niche Management approach and the Multi-Level Perspective to a case study in a developing country. And thirdly, it is the first paper that includes two interacting regimes in the MLP analysis.
The paper is practically relevant for two reasons. Firstly, although prior to the launch of the NBPE comprehensive sources of information were developed [4,10], and, as also mentioned above, recently two papers [8,9] described the status of biogas in Ethiopia and Sub-Saharan Africa respectively, based on findings on written data from 2011 and older, currently no up-to-date information or official statements have been published regarding the current status of development of domestic biogas or the drivers and barriers related to deploying it. Secondly, as similar biogas programmes have been promoted by SNV Development Organisation in Asia, Africa and recently in Latin America [11], insights in current status and barriers and drivers in other countries may yield useful insights for these new programmes.
The paper is structured as follows. Section 2 provides some technical background about the domestic biodigester. In Section 3, we discuss the key notions of the analytical framework applied. Section 3 presents methodology. Section 4 describes the development and current status of the Ethiopian biogas niche. Subsequently, the important factors and dynamics within the niche are investigated in more depth in Section 5 by using Strategic Niche Management. In Sections 6 and 7 we investigate the important factors and dynamics from outside the niche in more depth by using the Multi-Level Perspective. Section 8 presents our conclusions, theoretical findings and recommendations for actors involved.

Technical backgrounddomestic biodigesters
Domestic biogas can be produced in domestic biogas plants, also known as domestic biodigesters. Domestic biodigesters consist of small-scale microbe-controlled semi-batch reactors that process the organic compounds that are fed thereby producing a mix of 50-70% methane (CH 4 ) and 30-40% carbon dioxide (CO 2 ), among other gases [7]. Domestic biodigesters vary in size and shape but the principles behind their operation are similar. In the inlet, a one to one mix of manure and water is fed on a daily basis. The digester consists of a sealed structure that hosts the microbial activity and yields the biogas. The resulting bio-slurry leaves the biodigester through an outlet and can be stored in a pit, installed next to the biodigester. The most developed domestic biodigester technology in Africa and Asia to date is the fixed-dome digester [4]. The fixed-dome digester consists of a stationary underground structure made out of cement, bricks or stones, sand and aggregates. The biogas piping system can be constructed with PVC pipes, flexible hosepipes or metal pipes. Fig. 1 depicts an example of a fixed-dome biogas digester.
Based on a daily manure production from four cattle heads 1 , domestic biogas can replace the equivalent consumption of five kilograms of firewood, 1.5 kg of charcoal or 0.6 l of kerosene per day [4,6]. The biogas that is produced is mainly used for cooking. However, it can also be used in biogas lamps. Biogas can also be used to power internal combustion engines, refrigerators or radiant heaters; yet their application is even less widespread as lighting or cooking [12]. Domestic biodigesters can raise the use of cleaner energy sources in Ethiopia and in parallel offer valuable co-benefits to their users such as increased agricultural productivity from the use of bio-slurry as fertilizer, avoided deforestation, climate change mitigation, reduced workload and time savings through the avoidance of firewood collection and reduced indoor air pollution.

Analytical framework
The analytical framework used in this paper is based on two socio-technical approaches: Strategic Niche Management (SNM) and the Multi-Level Perspective (MLP). With the term sociotechnical we mean that the topic under study contains both technical and social elements, which are interlinked. SNM is a theoretical framework that can be used to study the sociotechnical dynamics and factors within a niche around a new innovation [14,15] in this case domestic biodigesters in Ethiopia. The MLP adds to the SNM framework by giving insight into the external environment in which the new innovation is developingin this case the energy sector and the agriculture sector in Ethiopia and the Ethiopian society as a whole. The MLP approach studies how innovation is influenced by factors at three levels: the exogenous 'landscape'; the dominant way of providing a societal function or the 'regime'; and the 'niche', the level where the innovation emerges and develops [16]. This framework provides more insight into which factors are relevant, and how they interact.

Strategic Niche Management
Strategic Niche Management (SNM) was developed as an analytical approach that can be used to review and analyse the development of innovative technologies in niches, which can be seen as incubation rooms or protective systems surrounding the new technology [17][18][19].
In the niche, the innovation can grow and develop to become viable through gradual experimentation and learning by networks of actors. During this period, the emerging technology has to compete with the existing technologies which are technologically and economically superior to it [16]. These established technologies are part of large social networks, the regimes, which have certain rules such as price/performance ratio, engineering practices, user preferences and regulatory requirements.
In the initial stages, a niche technology finds itself within a technological niche, which is a space protected from the rules of  [5,13]. the regime, e.g. by subsidies or regulatory exemptions. A technological niche can evolve into a market niche, a space where users start to recognise the values of the innovation and where it is able to compete over the established technologies. Market niches can eventually lead to the development of a new regime or become part of it [14,17].
To analyse the development of a niche, researchers have proposed three niche processes which are dynamically interrelated: the voicing and shaping of expectations, network formation and learning processes [14]. We discuss them subsequently.

Dynamics of expectations
Expectations give direction to the technology development, influence design choices, and attract resources as well as new actors. According to [20], expectations contribute to successful niche development if they become more robust (shared by more actors), more specific (give guidance) and have a higher quality (the expectations are validated by the actual developments). In the early niche stages, participants join the niche by investing effort, money and time because they have expectations of the future success. At that moment, actors have broad and unclear expectations about the technology and different visions of its future [15]. In time, expectations can change because of external factors (regime and landscape) and internal circumstances (e.g. results from experiments within the niche) [14].

Network formation
Actor networks are essential for niche development since they sustain development, attract resources and new actors, enable learning and carry expectations [15]. Two characteristics are important when analysing the actor network. First, the network composition is an essential factor. A good network requires a heterogeneous group of actors with different interests and roles [14]. Secondly, the network should be aligned. This characteristic refers to the degree to which actors' visions, expectations and strategies are in line with the niche development. This alignment can be achieved through regular interaction and co-operation between the different actors [15].

Learning processes
Learning influences the niche by affecting the expectations and aligning them. A good learning process is reflexive and focuses on many aspects [15]. Furthermore, good learning processes should not be confined to individual learning by actors, but should also consist of interactive learning or, in other words, knowledge sharing among actors [21]. Interactive learning can be facilitated by, among other things, trust and proximity between actors or intermediary actors such as umbrella organisations which facilitate knowledge flows between other actors [21][22][23][24]. Hoogma et al. (2002) gives two types of learning: first order learning and second order learning. First order learning refers to learning about the innovation's effectiveness in achieving pre-defined goals. It is directed on gathering facts and data. Second order learning is learning about the underlying norms and values related to the new technology. This type of learning enables changes in assumptions, approaches and cognitive frames, and has a larger contribution to niche development than first order learning [18]. Furthermore, Hoogma et al. (2002) distinguishes learning with regard to the following five aspects: technical development and infrastructure, industrial development, social and environmental impact, development of the user context and government policy and regulatory framework.

Summary of niche indicators
Summarizing, we analyse the three niche processes by evaluating the indicators shown in Table 1.

Multi-Level Perspective
The upscaling of an innovation is not solely the result of the above described internal niche dynamics, the external environment also exerts influence. The Multi-Level Perspective adds to Strategic Niche Management to analyse the major external developments that affect niche upscaling. It divides the sociotechnical system into three different levels: the socio-technical landscape (macro-level), the socio-technical regime (meso-level) and the niche (micro-level) [25,26].

Landscape
The landscape consists of the deep structural trends and factors that are not part of the regime and niche, but influence them. A broad range of factors and processes can be taken into account at this level, amongst others macro-economic factors (e.g. oil prices, economic growth), population growth, level of corruption, cultural aspects such as status, power differences and presence of different tribes or classes; and availability of raw materials [26,27]. The landscape level has the slowest dynamics; these trends usually change relatively slowly and are hard to change. However, this level also includes unexpected events within or outside the country such as wars and oil price fluctuations [16]. In this paper we also consider funding programmes or technology programmes from abroad as landscape factors.

Regime
Situated below the landscape, the socio-technical regime is the level of the established technologies. The regime itself is generally stable and there is commonly resistance to the introduction of new technologies. This is because existing technologies are 'locked-in' or path dependent [28].

Niche
In niches, new technologies are developed, often to solve problems in the dominant regime. Because of its weak structuration (low stability and high uncertainties), the niche can easily be influenced by the regime and landscape [16]. Often, more than one niche are in development at the same time.

Interactions
The extent to which an innovation is able to upscale is influenced by the interaction between the three levels (landscape, regime and niche). As mentioned above, the room for niches in the regime is directly related to the stability within the regime. The actors in the dominant regime generally have an aversion against niche developments; the more stable the regime, the stronger the resistance for new technologies. A destabilised or weak regime offers windows of opportunity for niche breakthrough. Regime destabilisation originates from pressurising landscape factors and internal regime tensions. A niche can develop internal momentum through improved price/performance, support from powerful  [16].
When the interactions between the levels are aligned they reinforce each other. This process depends on the timing and the nature of multi-level interactions. This offers windows of opportunity for the radical innovation at niche level to break through in the dominant regime [16].

Summary of MLP indicators
Summarizing, we analyse the three niche processes by evaluating the indicators shown in Table 2.

Using the MLP framework for the analysis
The analysis of the development of domestic biogas in Ethiopia was divided in its socio-technical levels niche, regime and landscape. Strategic Niche Management (SNM) was applied to the case study material related to the biogas sector to give a systematic overview of the niche level where domestic biogas technology is being deployed. The landscape level and the relevant regime levels within the case study material (the broader energy and agriculture sector and the situation in Ethiopia as a whole) were assessed as external socio-technical levels that influence the development in the niche. The analysis of landscape, regimes and niche provided specific insights that allowed reconstructing the current development of domestic biogas and barriers and drivers for its further growth, which are stressed in Sections 5 and 6 in concluding subsections at the end of the analysis of each level. But first, Section 4 provides a description of the development of domestic biogas in Ethiopia.

Introduction and early development of domestic biogas
The early years of implementation of biogas in Ethiopia were not solely focused on small-scale domestic biogas; institutional and large-scale digesters were also deployed. A common characteristic was that the experiments were conducted on an isolated manner without proper means to up-scale the technology. Biogas was first introduced in Ethiopia by Ambo Agricultural College around 1957 to supply the energy for welding agricultural tools. During the 1970s, two biogas plants were introduced by the Food and Agriculture Organisation (FAO) as pilot projects to promote the technology [29]. During the last two decades, around 1000 biogas plants were deployed in Ethiopia with sizes ranging between 2.5 and 200 m 3 for households, communities and institutions [13]. During this period, different models were used (e.g. fixed-dome, Indian floating-drum and bag digesters). However, according to multiple consulted actors there was no local capacity to neither up-scale the technology nor sustain it. Hence, just 40% of the aforementioned biodigesters are still operational [10].
Between 1999 and 2002, Christopher Kellner, a German biogas expert, built 60 fixed-dome biogas plants through a bottom-up implementation approach he titled 'From the Point to the Area'. This deployment was partly done based on an Ethiopian-German development project titled 'Land Use Planning and Resource Management, Oromia' (LUPO). His implementation method consisted of the construction of an initial biogas plant and the subsequent promotion within the close surroundings. This triggered local demand from neighbours and propitiated the construction of additional biogas plants. However, the rate of implementation was slow due to the limited size of the trained labour force (basically, Mr. Kellner and a technician he hired). In 2002, Mr. Kellner departed from Ethiopia and wrote a manual for the construction of LUPO digesters, which were adapted to the Ethiopian context [30,31].
Furthermore, since 2000, LEM-Ethiopia (a local NGO) started an awareness and promotion programme with latrine-fed biogas digester in schools and households in regions like Amhara, Oromia and SNNPR [32]. To date, 22 and 25 of these digesters have been installed in schools and households, respectively. However, no data was found to determine whether those plants are still operational.
In 2006, a technical team integrated by experts from the Ethiopian Rural Development and Promotion Centre (EREDPC) and SNV conducted a feasibility study to determine biogas potential in four Regions of Ethiopia which accounted for the largest number of inhabitants and livestock (Amhara, Oromia, SNNPR and Tigray). Table 3 presents a summary of the technical potential for domestic biogas in Ethiopia based on the aforementioned feasibility assessment. Two scenarios (low and high) were calculated based on the availability and access to water resources in each region. It was estimated that between 1.1 and 3.5 million households could benefit from the technology. Consequently, a 5-year pilot domestic biogas programme was proposed with an estimated cost of 11 million EU aiming to deploy 10,000 biogas plants over this period [10].
In 2007, Dutch experts conducted studies regarding technical potential of domestic biogas in several African countries, including Ethiopia. The rising interest in the technology led to the launch of 'Biogas for Better Life, An African Initiative' in Nairobi in May 2007. Together with the launch of this initiative a business plan was developed with the aim to construct two million biogas plants by 2020, create 800 private biogas companies and 200 biogas appliance-manufacturing companies. [33]

Deployment of the NBPE
Inspired by the positive environment with regard to domestic biogas in Africa and with the aim to up-scale domestic biogas in Ethiopia, the Ethiopian National Biogas Programme (NBPE) was developed and launched for a first stage of implementation between 2008 and 2013. From February to July 2007, a team from SNV and EREDPC conducted an extensive consultation process with relevant stakeholders in order to develop a Project Implementation Document or PID [4]. On June 16th, 2007 the draft of the PID was presented and approved. Even though the assessment report suggested constructing 10,000 domestic biogas plants over a 5-year programme, the Ethiopian Government decided to set a more ambitious target. Consequently it was agreed to build 14,000 family-sized biogas plants between 2008 and 2013 and the NBPE was launched [34].
Furthermore, a subsidy was provided to biogas users to compensate for the initial cost and hence improve the affordability of the biogas plants [35]. Moreover, the subsidy was provided to compensate for the 'free' availability and predominant use of firewood. The provision of the subsidy depended on the compliance of the technical criteria set by the NBPE. Hence it was limited to a single model for the biogas digester: the SINIDU model which is an adaptation of the Nepalese GGC-2047 fixed-dome digester [4]. This subsidy is still been granted nowadays [35].
The first years of implementation of the NBPE faced several obstacles in deploying the programme. Biogas was considered a new technology in the rural context and demand turned out to be lower than expected. In this period, SNV installed 98 demonstration biodigesters in the four regions for demonstration purposes and decided to expand the number of woredas where the NBPE could be implemented aiming to catalyse biogas dissemination [36].
The slow development of the NBPE was evidenced by the small amount of biogas plants that were built. Moreover, the situation was exacerbated by a cement crisis faced by Ethiopia between 2010 and 2011 [34,35]. Consequently, in 2010, during an intermediate revision of the NBPE, the African Biogas Partnership Programme (ABPP) decided to reduce the initial target from 14,000 to 10,000 biogas plants by end of the first phase in 2013 [35][36][37][38]. The first phase of the NBPE culminated on December 31st, 2013. Official sources have stated that 8063 biogas plants were built during this phase and distributed as follows: 2480 biogas plants in Oromia, 1992 in Tigray, 1892 in Amhara and 1699 in SNNPR [35].

Present and future of the NBPE
The second phase of the NBPE takes place from 2014 until 2017 and aims to construct 20,000 additional biogas plants. In 2014 alone, the government expected to build 3600 biodigesters in the four regions of implementation [35]. A significant goal for this second phase has been to enable private sector involvement, a goal that was not accomplished in the first stage of the NBPE [36,39,40]. Fig. 2 presents a graphical summary of the number of domestic biogas plants constructed in Ethiopia since their introduction in 1957. Also the change in target in the first phase from 14,000 to 10,000 biogas plants can be seen, as well as the target in the second phase to build 20,000 additional biogas plants. The current status of the biogas sector will be deeper looked into in the subsequent sections.

Analysis of the socio-technical niche
Biogas in Ethiopia is currently in the niche phase. It holds a small share of the market and it is sheltered by way of implementation programmes. However, besides biogas also other niches around cleaner energy technologies exist in Ethiopia nowadays. For example, a large variety of improved cook stoves are being commercialised with the help of local actors such as the Horn of Africa Regional Environmental Centre & Network (HOA-REC&N) and GIZ-Ethiopia. Moreover, other renewables, especially picosolar is starting to spread in the country and private enterprises are beginning to appear for the installation and commercialisation of the technology [41]. Furthermore, large-scale power plants, like the Grand Ethiopian Renaissance hydro project and Corbetti Geothermal also stand out as new innovations within the energy sector [42]. So, the development of domestic biogas is not an isolated effort. However, as the markets for part of these niches are still small (pico-solar, improved cook stoves) or because they aim for a different market (grid based electricity for the cases of hydro energy and geothermal energy), they are not in competition with biogas at this moment. Therefore, in this paper we will not look further into these other niches and their interaction with biogas. In the remainder of this section we will analyse the biogas niche in Ethiopia by focusing on the niche processes and indicators that we presented in Table 1 in Section 2.

Network formation
Network formation looks into the composition of the network and alignment of the actors within it in order to assess their influence on the development of the niche [14,43]. In the case of Ethiopia, the NBPE defined the actors that should be present in the development of the programme and assigned specific roles to each one of them [4]. Moreover, it divided actors based on the different jurisdictional domains (federal, regional and local) and their respective backgrounds: government, NGOS, Micro-Finance Institutions, private sector, biogas users, etc.
Currently, the network of actors within the Ethiopian biogas niche is diverse and their alignment is poor. The current core of the biogas niche is formed by the NBPE. However complementing efforts and independent initiatives are also present and aim to fill the gaps left by the NBPE. Although cooperation linkages occur among actors, in many cases cooperation is weak or non-existent.  In the following sub-sections, the alignment among stakeholders is revised at three different layers of implementation: (1) national and international, (2) regional and (3) local.

National and international level
At the federal level, the NBPE is led by the National Biogas Programme Coordination Office (NBPCO), which is hosted by the Ministry of Water, Irrigation and Energy (MoWE). The National Biogas Sector Steering Committee (NBSSC) was proposed since the implementation of the NBPE and allows integration of actors from the Ministry of Agriculture (MoA), Ministry of Finance and Economic Development (MoFED), Ministry of Environmental Protection and Forestry, as well a representatives from the private sector and NGOs [4,35]. However, weak linkages between the Ministry of Water, Irrigation and Energy and the Ministry of Agriculture undermines inter-ministerial cooperation [44].
The Ministry of Agriculture manages an extensive database of farmers, as well as the number and type of livestock they own [44], a resource that would be of high value to identify potential customers. Furthermore, the Ministry of Agriculture has a direct line of command with its extension officers who work at a local level and have direct contact with farmers. Meanwhile, the complex structure of the NBPE results in redundant hierarchies and bureaucracy [39]. The fact that the Ministry of Agriculture does not play an active role in the implementation of the NBPE creates an institutional schism for the dissemination of domestic biogas.
At an international level, most of the stakeholders provide technical assistance and funding to the NBPE or other domestic biogas initiatives in Ethiopia. The funding for the NBPE is channelled by the African Biogas Partnership Programme (ABPP), based in Kenya. The ABPP allocates resources for each of the national biogas programmes it manages based on their yearly performance [39].
There is a strong consensus regarding the fact that the Government of Ethiopia is determined to play a strong role in domestic biogas disseminationas well as other sustainable energy technologies - [34,39,41]. However, this strong presence of the Government hinders private entrepreneurship and the lack of institutional capacity results in a slow rate of implementation of the technology [39]. Furthermore, even though efforts have been made by international actors to improve the decision-making and the planning process, the actual implementation of those improved mechanisms does not happen.

Regional level
Although implementation is carried out at a local level, most of the activities are managed and supervised at a regional level. Most of the stakeholders are grouped at this level (as can be seen in the stakeholder map in Fig. 3). The majority of the Universities operate at a regional level. The private sector related to domestic biogas is almost non-existent and most of the implementation is done by individual masons trained by the NBPE at the local level except from SELAM Group, a local private manufacturer and provider of most biogas stoves to the NBPE that is present in Addis Ababa and Awassa. Only two NGOs working with domestic biogas were found in Ethiopia. These are LEM-Ethiopia and the Institute for Sustainable Development (ISD) who work on promotion and awareness, and user training on bio-slurry utilisation, respectively. Furthermore, SNV Development Organisation plays a role as technical advisor in the implementation and is a continuous promoter of the programme since its beginning.

Local level
The implementation of domestic biogas happens at the local level and is usually managed and promoted at woreda (district) level. The Mines and Energy Agency (MEA) 2 appoints Energy Experts at each of the active woredas who are in charge for the training and supervision of the masons [35]. However, a hierarchy issue between the Regional Biogas Programme Coordination Offices (RBPCO)appointed by the NBPEand the Woreda Administration undermines the availability of the Energy Experts. The latter assigns the tasks to be conducted by the Energy Experts, which are not exclusive to biogas [39]. This causes a constant negotiation between the Regional Biogas Partnership Coordination Offices (RBPCO) and the Woreda Administration and undermines implementation.
The masons are rarely grouped as local micro-enterprises, however there is consensus at the regional and federal level to leverage entrepreneurship. Furthermore, it is difficult to maintain trained masons within the NBPE, due to the fact that they do not work exclusively on biodigester construction and are usually attracted to other construction activities where they can find larger profit margins [36,45].
As explained before, even though extension workers from the Ministry of the Agriculture are present at the local level and work directly with the farmers, there is no linkage at this level between the NBPE and the extension workers [39]. Consequently, the masons or the Energy Experts usually train biogas users on bioslurry utilisation although the focus and expertise of this human resource is not agriculturally related [44].

Barriers and drivers at the niche levelnetwork formation
All in all, the NBPE set a triggering institutional framework for the dissemination of domestic biogas in Ethiopia. Moreover, the programme identified and assigned roles and responsibilities for a diverse group of actors at the different jurisdictional domains. According to [14], the diversification of actors at the niche level is desirable to promote the development of the niche. Furthermore, according to [43] it is important for the network of actors to integrate stakeholders from the dominant regimes in order to increase the chances for the radical innovation to emerge. In the case of domestic biogas in Ethiopia, the NBPE set structures where actors from other governmental agencies could interact with the development of the programme. Although the composition of the network of actors is diverse, the alignment between the actors is weak. Moreover, the highly centralized and hierarchical nature of the programme hinders the contribution of an important actor type, namely the private sector.

Development of expectations
The launch of the National Biogas Programme in Ethiopia (NBPE) was a catalyser for the development of domestic biogas in the country. Prior to the NBPE, the development was scarce and limited to isolated efforts [46]. The introduction of a systematic approach to disseminate the technology allowed the integration of several stakeholders and disrupted the inertia imposed by the dominant regime with a predominant use of firewood. Nevertheless, the ambitions of the NBPE had to face and adjust to the constraints and challenges of the local context. Now, after the culmination of the first phase of implementation of the NBPE and in the first years of the second phase it is possible to compare and analyse how expectations have been shaped over this period. Table 4 summarises some of the main targets set in the project implementation document of the NBPE before the start of the first phase and it is compared with the results after its completion. Given the nature of the NBPE, a strong interest has been focused on the number of biogas plants that can be deployed by the programme [30].
Despite the initial enthusiasm, the NBPE promoters had to transform the promises and words written in the Project Implementation Document (PID) into real implementation structures, which meant: hiring staff, setting offices, training technicians and (most importantly) introducing a technology that was relatively novel to the country. At the same time, it was expected for the NBPE to start generating concrete results. Fig. 4 shows the number of biogas plants that were built each year during the first phase of the NBPE. It clearly shows that the first years of the NBPE could not match the expectations.
Although the NBPE faced a harsh environment, it was decided to extend the number of active woredas (districts) with the aim to increase the coverage of the programme and hence, reach more potential clients [36]. By 2010, the NBPE gradually started to take off but it encountered another barrier: between 2010 and 2011, Ethiopia faced a cement crisis [35].
From an outside perspective, it was difficult to appreciate the local efforts of the main promoters of the NBPE. Fund allocation by international donors is based on the periodical revision of NBPE's performance which is based on the number of biogas plants that are constructed. Actual implementation efforts were not able to comply with the targets. This mismatch deteriorated the overall expectations of the stakeholders involved and in 2010 (halfway through the development of the first phase), a silent decision was taken by the African Biogas Partnership Programme (ABPP) to reduce the initial target to 10,000 biogas digesters [35,36]. This decision shows how the expectations about the programme were significantly reduced.
From 2011 onwards, the NBPE started to deliver better results (as shown in Fig. 4); however, it was still not possible to meet the yearly targets. After the completion of the first phase of the NBPE, the programme was able to deploy 58% of its initial target (8063 biogas plants out of 14,000). Despite this bad performance, the NBPE significantly increased its area of coverage to 102 more woredas than the 28 woredas that were initially targeted [47].
To date, another unaccomplished target has been the empowerment of local Biogas Construction Enterprises (BCEs) [35]. Several factors have influenced a poor private sector involvement but what is worrying is that so far, the NBPE has not been able to find a solution for this issue [36]. Nowadays and especially for the implementation of the second phase of the NBPE, there is a general consensus among stakeholders that private sector involvement should actually take off during this period. But our interviews demonstrated neither a concrete understanding why this happens nor a plan on how to approach it [34-36,39]. One Table 4 Targets and accomplishments of the 1st phase of the NBPE [34,35].

Indicator
Initial target Actual implementation

Barriers and drivers at the niche leveldynamics of expectations
According to [43], expectations can change in strength, quality or realism. In the case of Ethiopia the inability of the NBPE to deliver what was initially intended forced the actors to reduce their expectations according to the actual implementation in order for them to become more realistic. Nevertheless, it is paradoxical that although this change was performed for the first phase of the NBPE, the current (second) phase of implementation has an even more ambitious goal to deploy 20,000 additional biogas plants by 2017. Fig. 4 shows that the deployment during the last years of implementation of the first phase increased significantly. However, it is questionable whether the new expectations will be able to meet the actual implementation or will fail to become unrealistic again. Furthermore, analysing the performance solely based on the number of biogas plants may distract from an important fact that is whether the biogas plants that are being deployed actually meet the needs of its users or if they remain operational after they are deployed.

Learning processes
After the development of the feasibility assessment in 2006, it was revealed that most of the by then existing biogas plants were not operational and the majority of them had been dismantled. Hence, a set of issues was identified and discussed so that it could be taken into consideration in the design of implementation of the NBPE [10]. Table 5 presents a set of lessons learnt based upon the aforementioned issues. They are divided into three categories as also distinguished by Hoogma (2002), as described in Section 2 of this paper: technical development, development of the user context, and government and regulatory framework. In the following sub-sections, an analysis of the current progress on the main issues from Table 5 is provided.

Technical development
To date, domestic biogas systems are unable to attend a comprehensive share of a households' energy demand. Yet most worryingly, injera baking cannot be pursued using biogas since it does not provide sufficient energy to perform this task [34,38,49]. As will be explained in Section 7.1, traditionally 50-60% of household energy demand is used to bake injera; a traditional fermented flatbread with a sour taste [2,49]. Although recent innovations have been developed to make injera baking possible using biogas, these innovations have not yet reached the commercial stage [35,38].
Furthermore, although the cement crisis from 2010/11 has been overcome [35,36], other components of the biogas plants are often unavailable in the local markets or are prone to malfunction. This issue is exacerbated by the fact that these problematic components, specifically the biogas valves and the biogas lamp, are imported and have a poor quality [38,45]. Moreover, the biogas stoves manufactured locally by SELAM Group are scarce and their price and quality fluctuates considerably [38].
Moreover, the role of after-sales service is key to assure the operability of the plants over their lifetimes. However, the only mechanism for quality assurance after the construction of the biogas plants is done through a sampling process by the Regional  [39,47,48]).

Table 5
Biogas issues identified by the NBPE's feasibility analysis in the feasibility study of the NBPE in 2006 [10].

Category Issue
Technical development Domestic biogas installations cannot supply the full domestic energy demand Non-local materials increase investment costs and maintenance problems In relation with the available dung, most installations are over-sized Without proper technical back-up, any plant will fail sooner or later 'Single actor construction' weakens local technical back-up facility Standardisation will improve quality Development of the user context Farmers need proper instruction to maximise the benefits from their investment Biogas installations as a 'stand-alone' application are likely to fail The plant's water requirement shall not be underestimated Government context and regulatory framework EREDPC seems the best placed lead agency for a federal domestic biogas programme The Bureaus of Energy would seem the best placed coordinating, supervising and integrating agencies at regional level The extension network of the Bureaus of Agriculture would be an asset for a large scale dissemination programme The regional micro-finance institutions can play an important role in a large scale domestic biogas programme Biogas Technicians. When asked about the percentage of biodigesters deployed that are still operational, the answer was that the NBPE have not pursued a biogas inventory yet and there was no definite numbers on this topic [45]. However, during our field study in Ethiopia some interviewees mentioned that a large percentage of biodigesters were malfunctioning or out of operation. According to the NBPE promoters and especially SNV worldwide, one of the main drivers for biogas dissemination is the systematic approach which includes involving a broad range of stakeholders [34]. The NBPE was able to transition from a 'single actor' approach to disseminate biogas technology to a multistakeholder approach [13]. However, as described in subsection 5.3.1, this has not led to a significant increase of local technical backup facilities and the rate of operability of the biogas plants that are already installed is questionable.
Furthermore, the standardisation of the SINIDU fixed-dome model has eased the quality assurance procedures. Nevertheless, similarly to what happened in Nepal during its initial phases of implementation, at some point it became necessary to diversify the technologies for domestic biogas production [50]. Also, in Ethiopia isolated efforts outside the NBPE have started to appear trying to address the gaps or burdens of this single-technology approach by diversifying the models and business plans to disseminate domestic biogas [38,40,41]. However, it is still uncertain if these complementing initiatives will be considered by the NBPE [35]; which was eventually the case in Nepal, where it was decided to diversify the technological options for domestic biogas [51].

Development of user context
According to [30], it is common that due to lack of knowledge and awareness, many of the biogas users do not exploit the full potential of biogas plants. For example, bad habits such as irregular or insufficient feeding of the biodigester significantly hinder biogas production. Further, as reported by Araya (2014), bio-slurry utilisation is poorly taught to biogas users by masons and although extension officers from the Ministry of Agriculture could perform this task, it is not done due to abovementioned inter-ministerial misalignments A significant challenge in many areas of Ethiopia is water scarcity and drought during certain times of the year. As explained in Section 1, under the traditional fixed-dome model an equal amount of manure and water has to be provided daily [7]. In order to remediate this issue, around 50% of the biodigesters that are being deployed include a toilet connection, which enables the use of urine to compensate for the lack of water [45]. However this contribution is negligible in comparison to the daily amount of excretes or liquid that are required [5]. Currently, some member countries of the African Biogas Partnership Programme (ABPP) are experimenting with Solid-State Digesters (SSD) [52]. The SSD is a modification of the conventional fixed-dome model and allows using a 4:1 manure/water mix instead of 1:1, hence significantly reducing the water demand of the biodigester (SNV, 2013b). However, SSD has not yet been introduced to Ethiopia [35,36,39].

Government context and regulatory framework
Although initially the Ethiopian Rural Development and Promotion Centre (EREDPC) was appointed as the leading organisation at the national level, this organisation ceased to exist due to a ministerial restructuration [4,35]. At the federal level, the whole staff was substituted. However, labour conditions did not change nor improved compared to other similar government positions (lower salaries, no pension fund, no health care, etc.). This incurred in high personnel turnover and unfulfilled vacancies [53].
No official data was found specifying the details of the transition of the NBPE from the former Ministry of Mines and Energy to the Ministry of Water, Irrigation and Energy (MoWE). Therefore, it remains uncertain if elements such as institutional capacity, human resources or allocated budget were altered due to this transition. On the other hand, the abovementioned institutional transformation did not affect the regional level, and the Bureaus of Water, Energy and Mines (MEAs) remained unaltered during this ministerial restructuration [35,45]. What is certain is that, at the federal level, labour conditions neither changed nor improved: the staff remained with low salaries, no pension fund nor health careconditions that other government officials do have. This situation incurred in a high personnel turnover and unfulfilled vacancies [53].
Furthermore, in Table 5, it is interesting to see that prior to launch of the NBPE it was recommended to incorporate the extension officers from the Ministry of Agriculture to provide support in the implementation of the programme [10]. However, as explained before, the incorporation of the extension officers has not happened yet due to inter-ministerial misalignments between the Ministry of Agriculture and MoWE [44]. On the bright side, micro-finance institutions have been gradually integrated within the NBPE and around 57% of the biogas plants that were constructed by the NBPE used micro-loans as a means to fund the initial investment. Moreover, the World Bank has allocated economic resources to improve micro-credit access for renewables, including biogas. These funds are being channelled by the Development Bank of Ethiopia and distributed to the regional and local micro-finance institutions [35]. To date, two Micro-Finance Institutions have been able to attain this capital loan [53].

Barriers and drivers at the niche levellearning
All in all, from a learning processes perspective, the NBPE was able to attain first order learning among NBPE actors and governmental actors; however, it was generally unsuccessful to generate second order learning. In other words, initially NBPE was able to determine several factors that could poses a risk for the implementation of the programme (hence creating first degree learning). Nevertheless, it was unable to effectively solve the issues it initially identified by changing its basic assumptions and approach. Furthermore, first order learning among masons, technicians and users has not been optimal because of unclarity about who is responsible for this training and problems with expertise among trainers, as explained in subsection 5.1.3, and high illiteracy rates among the population.

Analysis of the socio-technical landscape
At the landscape level, economic instability and poverty in Ethiopia constitute the most relevant barriers for biogas dissemination. The Federal Democratic Republic of Ethiopia is categorised as one of the poorest countries in the world, although the country has experienced a sustained economic growth over the last years and the gross national income (GNI) per capita has more than doubled over the last 20 years. Poverty goes beyond per capita income; it affects people's quality of life and their ability to overcome their misery, hence poverty itself becomes a trap [54]. Because of this widespread poverty, many people in Ethiopia do not have enough purchasing power to buy a biogas plant. Ethiopia has the second largest population in Africa after Nigeria. Moreover its population has followed an almost continuous trend of growth of 2.2% per year over the last 10 years. By 2011, 83.2% of the Ethiopian inhabitants resided in rural areas and scattered over a 1.1 million km 2 territory [55].
Political instability constitutes an additional constraint, both internal and external to the country. From an internal perspective, although the country deposed the communist regime in the early 1990s and instituted a multi-party democracy there are still shadows from authoritarianism which are still present nowadays. According to [56], despite the sustained economic growth, the current enormous public intervention overshadows and even hinders private sector investments. Other factors that hinder entrepreneurial activities are inadequate financing possibilities, a shortage of educated and skilled labour [57], and the inability of firms to convert part of their profit into investment [58]. Although the Ethiopian government launched market reforms in 1991, the country remains in a vicious cycle of low productivity, low-paid jobs, and poverty [57]. From an external perspective, Ethiopia faces significant geopolitical issues since there are latent border conflicts with Eritrea and Somalia and wars in its neighbouring countries (e.g. Sudan and South Sudan, Somalia) that pose a constant threat to the political and economic stability of the country. The political and economic instability lead to a low overall level of competitiveness of the country. Nevertheless, the Ethiopian Government has ambitions and commitment to provide solutions for poverty eradication [56].
Another barrier is the low literacy rate in Ethiopia of 40% [55]. This makes the training of masons, technicians and users more difficult. This problem is exacerbated by the fact that within Ethiopia more than 80 different mother tongues exist. Although Amharr is the first language and many people speak English, in rural areas many people only speak their mother tongue (CSA, 2010). Another barrier is the lack of basic infrastructure which is evidenced by, among other things, the low access rates to clean water and household sanitation. The availability and access to water can represent a barrier for the deployment of the technology [59]. The appropriate operation of biogas digesters can be threatened due to seasonal or geographical shortages of water. Nevertheless, according to [60], and as described in Section 5 of this paper, the shortage of water can be complemented with the use of animal or human urine.
On the other hand, at the landscape level, there are also drivers that motivate the development of domestic biogas in Ethiopia. For example, the physical geography of Ethiopia complies with technical criteria for the biogas plants that are being deployed to operate in terms of temperature and availability of waste organic matter. According to [38] this suitability of the Ethiopian climate is not being fully exploited since the main model for the biogas plants is situated underground where there is a lower temperature than in the surface. Hence, the microbial activity inside the biodigester could improve with a rise in the temperature of the digester chamber if it would be located above the surface [31], hence causing an increase of the biogas yield. In addition, other drivers at the landscape level are scarcity of firewood, deforestation, depletion of soil nutrients and erosion [53]. The systematic depletion of the forest coverage to satisfy the energy needs has led to a systematic depletion of the forest coverage over the last 35 years and a shortage of traditional biomass resources. Nowadays, only 2.7% of the land is covered with forests [4]. The main driver at landscape level for the growth of the biogas sector in Ethiopia is a driver that originates from outside the country: the NBPE programme.

Energy regime
The Ethiopian energy regime is mainly based on biomass. According to [61], 92.9% of the primary energy consumption comes from biofuels and waste; 81.2% of which is supplied by woody biomass (especially firewood), followed by dung cakes and crop residue with a contribution of 9.1 and 8.1%, respectively. Other types of biomass are charcoal, bagasse and bio-briquettes [3]. Electricity provides only 1.1% of the energy demand and it is mostly supplied by hydroelectric plants: 93.9% of the electricity comes from hydro and the remaining is supplied by geothermal and fossil fuel-based power plants. According to [1], electricity access reached only 23% of the Ethiopian population by 2010. One of the problems resulting from Ethiopia's limited electricity use is the increase in health problems due to smoke emissions resulting from indoor cooking and indoor use of kerosene lamps. Moreover, it is estimated that in rural areas, women and children spend at least 5-6 h collecting fuel wood [53].
When looking at the demand side, the residential sector consumes 93.5% of the energy that is supplied [55]. The rest of the sectors have a negligible contribution to the energy demand, including industry with 1.8% and the commercial sector with 1.1% [61]. At the household level, 98.6% of the energy is supplied by biomass. According to [2,49], 50-60% of household energy demand is used for baking injera; a traditional fermented flatbread with a sour taste. Injera is baked on large batches using a clay plate covered with a lid made out of straw and dried cow manure [62]. The baking process is highly inefficient and uses a significant amount of firewood [10]. The remainder of the consumed energy is used for cooking other foods and for lighting.

Agriculture regime
Agriculture represents the main economic activity in Ethiopia, accounting for 48.8% of the country's GDP in 2012 [1]. Moreover, Ethiopia is believed to have one of the largest livestock populations in Africa [63] and it is globally recognised as the cradle for coffee, a deeply rooted beverage within its culture. The majority of the agricultural production in Ethiopia is performed at a household level, 96.8% of which takes place in a rural context, where crop harvesting and animal husbandry activities are often combined. The main products are cattle and cereals. In These agricultural households are characterised by limitedly small land tenure and it is common for all household members to contribute to the farm's activities.

Agricultural crops
The agricultural activity in Ethiopia is divided in two main seasons. The meher season, which relies on the precipitations from June to September and it is when most of the crops are harvested. And the belg season, which offers a second opportunity to a harvest a smaller batch of crops thanks to rainfall between March and April [64]. Furthermore, by far, the vast majority of Ethiopia's agricultural land is utilised for temporary crops, mainly to grow cereals. The injera mentioned in Section 6 is the traditional base of the Ethiopian diet and it consists on a fermented flatbread made out of a cereal named teff. Nowadays, due to the high prices of teff, it is very common to find injera made out from wheat [53].
Agricultural crops are concentrated in the regions of Amhara, Oromia, SNNPR and Tigray, which account for 97.4% of the total annual crop production. It is no coincidence that the Ethiopian National Biogas Programme (NBPE) takes places in these four regions, since most of the crop and livestock production takes place there.
Ethiopian farmers employ different practices in order to improve their crops' productivity. The most commonly used practice is to fertilizer application, mainly during the meher season (second half of the year). These synthetic fertilizers are applied to almost half of the cropland area, while natural fertilizers are only applied to 10.2% of this land. Moreover, irrigation alleviates water scarcity during the dry seasons between October and March. However, just 1.8% of the total cropland area employs irrigation practices [65].

Animal husbandry
Livestock activity plays a number of roles in Ethiopia's economy. It provides food, force, soil fertility and a distinctive means of savings at the farm level [66]. It represents a 'near cash' capital stock for the peasants which can be exchanged when the agricultural households face economic struggles [63]. As mentioned before, due to the scarcity of firewood in the country, many farmers have been forced to use manure as an alternative energy source through 'dung cakes' (sun-dried manure), causing soil depletion in exchange for an inefficient source of energy [5,67].
Similarly to crop production, most of the livestock is concentrated in the regions of Amhara, Oromia, SNNPR and Tigray. Oromia is the region with the highest cattle population with almost 22.5 million cattle heads. Camels are usually found in the Eastern and Northern regions of Ethiopia (e.g. Amhara, Tigray and Oromia), and are mostly used for dairy and transportation [63]. Other livestock present in Ethiopia are poultry and beehives. The total poultry population in Ethiopia accounted for 44.9 million by 2012, most of which were chicks (40%) and laying hens (33%). Meanwhile, apiculture represents a diversifying income for smallholder farmers and it represents an annual production of 39.7 million ton of honey and 3.8 million ton of beeswax. To put this in perspective, Ethiopia is the tenth largest producer of honey and the fourth largest producer of beeswax worldwide [68].

Barriers and drivers at the regime level
As also mentioned for the landscape level, the inability of the traditional practices to provide a solution for the current energy and environmental crisis constitute a driver for domestic biogas dissemination. The energy and agriculture regime are well tied in Ethiopia and interrelate with each other. The largest share of the energy consumption and most of the agricultural activities are conducted at the household level in the rural context [55,69,70].
Based on this motivation, the Ethiopian Government has focused on deploying actions to alleviate the energy crisis by providing national frameworks of implementation to embedded the implementation of alternative energy sources such as domestic biogas [71]. The creation of such frameworks has also promoted the creation of a suitable sectorial policy which is a driver for the development of alternative energy technologies [37].
However, albeit the unsustainability of the trends in the current dominant regimes such as the use of firewood a lock-in results in the perpetuation of traditional practices. Another barrier is the fact that current commercially available biodigesters are not suited for injera cooking, the traditional food in Ethiopia. Another important barrier in the regime is the poverty already mentioned in the landscape. The low purchasing power of Ethiopian farmers limits them to invest in a biogas plant. According to [39], an Ethiopian farmer faces the dilemma between investing in a biogas plant or increasing the number of cattle heads he or she poses. For the case of Ethiopia, a biogas digester costs the equivalent of three cows, whereas in Tanzania and Kenya it costs two and 1.5 cows, respectively.

Conclusions and recommendations
The current paper has looked into the development of domestic biogas in Ethiopia by analysing its current status as well as drivers and barriers for further growth. It has been found that the gradual emergence of the niche technology has been accomplished based on a systematic and centralized approach leaded by the Ethiopian National Biogas Programme (NBPE).
Our analysis shows that barriers and drivers for the deployment of domestic biogas in Ethiopia are present at the different sociotechnical levels. Moreover, there are strong interdependencies within and between the levels, e.g. between the energy and agriculture regimes. The main drivers and barriers we found are represented in Table 6.
As can be seen in this table, at the landscape level an important driver for the current growth of the biogas sector in Ethiopia is the NBPE programme, largely funded by the Dutch SNV Development Organisation. The deforestation resulting from the traditional use of firewood for cooking in Ethiopia is another driver, as is the climate which is very suitable for the operation of biodigesters. Barriers at the landscape level are the political and economic instability, the widespread poverty which results in insufficient Table 6 The main drivers and barriers for further growth of the biogas niche in Ethiopia.

Socio-technical level Barriers Drivers
Landscape Political and economic instability Low suitability of economic climate for enterprises and innovation Widespread povertytherefore most farmers have insufficient purchasing power to invest in a biogas plant Suitability of the climate Lack of natural resourceswater shortages Lack of natural resourcesdepletion of firewood High rate of illiteracy Presence of different tribes with different mother tongues Presence of NBPE programme Regime Stability in regimedeeply rooted practices of firewood use Suitability of sectoral policy Amount of lock-in in regimedeeply rooted practice of injera cooking which is currently not possible on biogas Amount of lock-in in regimedeeply rooted practice of firewood use Niche Not fully completeness of network of actorsprivate sector is missing Almost completeness of network of actors, including regime actors Poor alignment in network of actorsmany mismatches between governmental bodies and levels Mismatch between expectations and actual developmentimplementation goals have not been reached; private sector is still not involved Problems with first order learning among masons, technicians and users Presence of first order learning among NBPE actors and governmental actors Lack of second order learningissues present at start of NBPE programme are still not resolved purchasing power for most Ethiopians to invest in a biogas plant and the difficult circumstances for enterprises and innovation. Other barriers are the high illiteracy rate and the large number of mother tongues in Ethiopia which makes training of masons, technicians and users more difficult. In addition, water shortages may also become a barrier since a daily 1:1 ratio of manure and water needs to be fed into the biodigesters in order to assure there adequate operation.
As a consequence of the NBPE programme, suitable sectoral policies were set up at regime level, which are a driver for the further growth of biogas in Ethiopia. A barrier at this level is the deeply rooted practice of injera cooking, an Ethiopian staple food, which is not possible on the currently commercially available biogas cookers.
At the niche level, drivers are (1) the fact that many different actors are present in the network, among whom governmental actors at many different levels and also regime actors, and (2) the presence of first order learning about biogas technology, at least among NBPE actors and governmental actors. However, at this level also quite some barriers exist. Although many different actors participate, the private sector, which should play an important role, is currently not active in the Ethiopian biogas niche. Also, the alignment between the actors that are present in the niche is weak, especially between the different governmental bodies. Furthermore, because the actual implementation rate of biodigesters is severely lagging behind the initial goals, expectations had to be lowered during the recent years. Also, first order learning among masons, technicians and users has not been optimal because of problems with training and high illiteracy rates. Furthermore, second order learning is lackingissues that were already identified before the start of the NBPE programme, e.g. regarding technical problems, the user context and the regulatory framework, are still not resolved.
Our recommendation for actors involved is to focus quickly on addressing the abovementioned barriers, part of which have already been known for years. One aspect here form the emerging initiatives that are focused on solving the technical issues such as the inability of the biogas systems to satisfy a significant share of the household energy demand (in simple words, being able to bake injera with biogas). Another important aspect is the empowerment and inclusion of the private sector by rethinking the way the technology is disseminated. A third recommendation is to pay special attention to training and awareness of masons and users in the rural areas especially adapted to circumstances with high illiteracy rates and many different mother tongues. A fourth aspect to mention here is the price of the biodigesters and the price differences between different countries. In this respect, starting to disseminate different designs and sizes of biodigesters, as has also been done in Nepal, seems worth considering. These actions combined would improve the further dissemination of this technology and in that way contribute to innovation for sustainable growth.
From a scientific perspective, the combined application of Strategic Niche Management (SNM) and the Multi-Level Perspective (MLP) enabled us to derive deep and scientifically innovative insights into the status and dynamics of this case and the drivers and bottlenecks for further diffusion of biodigesters in Ethiopia. Also, it adds to the body of literature about innovation in the developing world. Moreover, the analysis of multiple relevant socio-technical regimes allowed to better understand the interrelationships between them as well as the influence that they exert on the niche technology.
A recommendation for further research which is also practically relevant is to also apply the framework used in this paper, including two regimes, to related case studies, such as biodigesters in Nepal and in Rwanda. Making a comparison between the drivers and barriers found in these cases and in the case of Ethiopia will be practically relevant and will also contribute to our insights into the impact of different contexts on the speed of the diffusion of the same technology. Another recommendation for further research is to further analyse the international aspects of this case and similar cases, where financial resources and technology are exported from one country (in this case The Netherlands) to another (in this case Ethiopia) while both the technology and the diffusion process are not optimally designed for the local context in the receiving country. This further research could build further on insights into international aspects of technological innovation systems as published in [72] and [73].