Data were collected from both primary and secondary sources. Secondary data were collected through literature review using published documents and internet material. There was also a review of policies related to energy in Tanzania. Secondary data helped to establish what has been done in the subject and to read what were the remained gaps for field work were. Institutions supporting biogas development were consulted for

understanding their performance and constraining factors they are facing. Primary data were collected in areas related to investment cost, awareness, household energy demand, technology service providers, and expertise. In addition, there were consultations with service providers to get information on cost, demand as well as factors constraining the spread of the biogas technology in the District. Furthermore, there was a consultation with local and district institutions and authorities for detailed information on biogas use in the district and whether there has been any efforts to facilitate the adoption of biogas.

The sample frame for this study involved respondents with dairy cattle/biogas use and those with dairy cattle but have not installed biogas plants. Also respondents with access to electricity and other clean energy sources such as LPG were included in the sample. A total of 3 villages were selected for the household sample. These were Isagilo, Kyimo and Mpandapanda. The selection of the villages based on the availability of dairy cows, adoption of biogas technology, availability of other energy sources, socio-economic status and accessibility. The households were selected purposively for those with biogas as well as those with access to electricity as they are few but random for the rest of the dairy keepers.

The total number of households (n) to be surveyed was estimated using the formula below:

A sample size of about 10% was selected making a total sample of 120 households. Out of this, 35 had biogas facilities and the remaining 85 had dairy cattle without biogas facility (Table 1). Village roster were used to select the sample households. Data were collected using structured and semi-structured questionnaires and analysed using Statistical Package for Social Sciences (SPSS) as well as livelihoods models. Results have been presented in tables and figures.