This paper develops a constraint-first framework for evaluating the long-term persistence of technological systems under thermodynamic limits. Beginning from finite energy flux, irreversible entropy production, component degradation, finite signal speed, and duration-weighted survival, it derives conditions under which maintenance burden scales superlinearly with energetic throughput. If maintenance scaling is superlinear, high-dissipation configurations become structurally fragile over gigayear timescales. Repeated selection suppresses variance in bulk energy-integrated observables while leaving orthogonal structural dimensions comparatively unconstrained. Increased sensitivity in a single energetic metric therefore does not guarantee improved separability after deep-time filtering. This work forms the thermodynamic layer of the Deep-Time Structural Constraints series.
Brenton Fournier (Mon,) studied this question.