Photovoltaic (PV)–building systems are commonly assessed using aggregated energy indicators that obscure the structural diversity of energy flows. This study introduces an entropy-based approach for quantifying the operational structure of energy flows in a real PV–building system from an information-theoretic perspective. High-resolution 15-min measurement data covering a full month were used to derive cumulative energy contributions associated with local self-consumption, surplus export to the grid, and electricity import from the grid. These flows were reformulated as a probability distribution and analyzed using Shannon entropy. The results reveal a high normalized entropy of monthly energy flows, indicating substantial dispersion among operating states despite the dominance of local self-consumption. Entropy decomposition shows that all flow components contribute meaningfully to system uncertainty, including surplus export, which exhibits a disproportionate informational impact relative to its energy share. The findings demonstrate that entropy captures structural properties of PV–building operation not accessible through conventional energy balance metrics. The proposed framework provides a compact, scale-independent descriptor of operational complexity and offers a new quantitative perspective for assessing predictability and structural heterogeneity in PV–building energy systems.
Arkadiusz Małek (Tue,) studied this question.