Explorative ex-ante consumer cluster delineation for electrification planning using image processing tools

Abstract

The ex-ante division of consumers into potential off-grid and grid-extension clusters represents a crucial input to electrification planning. This paper explores the application of image processing tools to distinguish grid-expansion from off-grid consumer clusters spatially, considering peak-load surfaces. Instead of point-based cluster analysis, the proposed model induces morphological pixel changes in the image through filtering and segmentation techniques. This way, load clusters can be isolated based on texture and coherence in an unprecedented way. Various single and sequential image processing techniques are compared, together with a rough grid expansion-ratio estimate for each case. Model outcomes are benchmarked against off-gridon-grid clusters retrieved by using the Rural Electrification Model.The ex-ante division of consumers into potential off-grid and grid-extension clusters represents a crucial input to electrification planning. This paper explores the application of image processing tools to distinguish grid-expansion from off-grid consumer clusters spatially, considering peak-load surfaces. Instead of point-based cluster analysis, the proposed model induces morphological pixel changes in the image through filtering and segmentation techniques. This way, load clusters can be isolated based on texture and coherence in an unprecedented way. Various single and sequential image processing techniques are compared, together with a rough grid expansion-ratio estimate for each case. Model outcomes are benchmarked against off-gridon-grid clusters retrieved by using the Rural Electrification Model.