In recent years, energy markets have been unstable, and the price volatility has been high. In addition, the EU has extended the Emission Trading System (ETS), which also affects energy prices. High energy prices may make low-income households vulnerable to energy poverty. The EU has set up a Social Climate Fund (SCF) to work towards a fair energy transition. SCF interventions require analysis of household energy poverty. This study analyzes data from Finnish low-income households to detect the extent to which there is a risk of energy poverty and to find the groups with the highest risk. Clustering algorithms in the k-means family are used, and fuzzy extensions of the methods are applied to account for overlap in features. Silhouette evaluation shows that fuzzy c-means clustering has the best ability to separate clusters from the data. Results reveal two main clusters: low-risk households in urban regions and high-risk households in rural regions. Housing- and transport-related energy poverty are both detected. One-third of the households exhibit a risk of energy poverty. The analysis contributes as an example of how analytics can be used for data-based public decision-making. The results are relevant for energy poverty-related decision-making and policy planning.