Clustering methodology for planning the implementation of pico-hydroelectric power plants on remote livestock farms in the Republic of Tuva

Abstract

A methodological approach to power supply for remote isolated livestock farms in the Republic of Tuva where free range animal husbandry lacks grid electrification is discussed. The use of pico-hydropower plants (up to 5 kW) operating on small watercourses is proposed. The research goal is to identify priority districts for this technology via multi-criteria analysis accounting for regional specificity. Four key criteria are developed: remoteness from grids, hydroelectric potential of small rivers and streams, farm energy consumption, and pasture seasonal patterns. Qualitative indicators are formalized into ordinal or binary scales after which hierarchical clustering (Manhattan metric, complete linkage) classifies 17 kozhuuns (districts). The cophenetic and silhouette coefficients confirm clustering adequacy. Four clusters with varying priority for pico-hydropower plant implementation are identified. Using an integral priority index (weighted sum of normalized criteria), all 17 districts are ranked. The Todzhinskiy and Mongun-Tayginskiy districts are high-priority areas where hydroelectric potential combines with maximum grid remoteness; pico-hydropower plant installation here is economically and socially justified. The second and third clusters require feasibility studies or seasonal solutions. For minimal priority districts (Erzinskiy, Ovyurskiy), pico-hydropower plants are impractical due to low hydroelectric potential or proximity to power lines. The proposed methodology enables objective territorial classification, differentiated energy supply approaches, and justification of investment priorities in decentralized renewable energy.