The pitfall of achieving sustainable energy for all in Kenya under climate impacts: Insights from time-series models
Ensuring access to reliable and sustainable energy (SDG 7) is critical for Kenya’s development, yet despite this importance, the fundamental drivers of its electricity demand are not well understood. Existing literature shows mixed results regarding the causal links between electricity consumption, population growth, and economic development. The dissertation addresses three key questions: (1) how population and GDP growth correlate with electricity consumption trends (1980-2023), (2) how climate change uncertainties will modify future consumption patterns to 2050, and (3) what policy strategies are required to sustain electrification progress.
Using an integrated modelling framework that combines Vector Autoregression (VAR) analysis with system dynamics forecasting, this research adopts a mixed-method approach that incorporates uncertainty analysis to ensure robust model validation and forecast reliability. It utilises annual time series data from 1980 to 2023 to analyse Kenya's electricity consumption patterns, generating projections to 2050 for distinct climate scenarios.
The analysis reveals that Kenya’s electricity consumption is predominantly driven by historical trends, not GDP or population growth. The autoregressive patterns show no statistically significant relationship to GDP growth or population growth, as indicated by Granger causality tests and forecast error variance decompositions. Forecasted projections for 2028 and 2050 remain moderate between 27.82 – 29.29 TWh, even under varying climate change scenarios(SSP1-2.6, SSP2-4.5 and SSP5-8.5), indicating a high degree of resilience to external shocks.
These findings provide evidence-based insights for policymakers, researchers and energy planners, demonstrating that energy planning must focus on internal system patterns rather than external economic pressures to build resilient infrastructure and achieve sustainable energy goals.