IN RUHIIRA:

Although households aren’t heavy electricity users, they are good indicators of energy requirements for an area. For instance, the denser the population, the more likely the need for other services like health facilities. Therefore, we started by mapping the population. Like all energy planning, we required a geographically located resolution of the population at 1 km² (ideally 500 m²) because 1 km is the maximum distance for reliable low voltage line operation.

We had existing polygon¹ population data, but to achieve the desired 1 km² resolution, we needed household data. If this data doesn’t already exist from census or other government sources, we generally gather this data in one of two ways. We either use satellite imagery to locate households or collect population data by household enumeration with Formhub, a tool we developed to make mobile data collection fast, flexible and easy.

In Ruhiira, because of the large area and poor roads, we decided against household enumeration. Instead, we created automated software (using image recognition in Python) for finding rooftops in high-resolution satellite imagery from QuickBird. Locating households using satellite imagery can also be done manually.

TIP: Household enumeration is a good option when existing site data is lacking and local governments are willing and able to use modernized data gathering tools. In these cases, focus on developing the capacity of local stakeholders for ongoing data acquisition and maintenance rather than collecting the data with your own team.

Our software identified 13,366 households in Ruhiira, which we clustered at a resolution of 500 m². In our model, we assumed the number of people in each household to be the same, which we found to be XX people per household.

< Prev | Next: Step 2 >