Characterization and prediction of household use of solar home system in Leyte Islands, Philippines
DOI:
https://doi.org/10.32945/atr4217.2020Keywords:
rural electrification, renewable energy system, nonparametric method, classification treesAbstract
Under the Household Electrification Program of the Philippines‘ Department of Energy, 530 solar home systems (SHSs) with 30Wp capacity were installed in individual households in four island barangays (villages) in Leyte and Southern Leyte from 2013 to 2015. Aside from lighting with four LED bulbs, the SHS can also be used to listen to the radio or MP3 player, and charge a mobile phone. This paper examines household use of the SHS thru data collected from 299 beneficiaries interviewed in the period 2017-2018. Results indicated that while all households use the SHS for lighting, its other applications are not utilized as much. A sizable percentage of households are now using only the SHS for lighting in each island barangay. The use of the generator set (for a limited time at night) in partnership with the SHS remains dominant to be able to use the SHS lights longer. Kerosene lamp is now rarely used for lighting except in the absence of a barangay-owned generator set.
With a substantial percentage of households using the SHS for lighting only amidst its other applications, classification trees by island barangay were generated to identify the variables which will explain SHS use with both quantitative and qualitative household characteristics making up the predictor space. Although the classification trees grown are unique to each island barangay, age of the household head and annual household income are the consistent significant predictors (up to the third most important) followed by median age of the household members.