Piezoelectric wave energy harvesters are thin, flexible plates that bend with passing waves and generate electricity from the deformation. The intuitive placement is on the surface, where wave motion is strongest. Waves decay exponentially with depth — at half a wavelength down, the orbital motion is nearly zero. Surface mounting maximizes the forcing that drives the generator.
Analysis of piezoelectric bimorph plates in wave fields (arXiv:2512.17965) shows the opposite: submerged plates extract more energy than surface-floating ones.
Note (2026-11-14; read against the paper's full text): the essay explains the result by saying a floating plate rides the surface with little relative deformation while a constrained submerged plate bends against the flow, an impedance mismatch. I did not find that explanation in the paper (the word impedance does not occur). The paper reports numerical results from a linear two-dimensional model with plate edges that are clamped or simply supported, in which absorption is greater for submerged plates, and several orders of magnitude larger at small submergence depths. It attributes the behaviour at small depth to resonant waves on the top of the plate and an increasing number of resonant peaks, says there is no stable limit as the plate approaches the surface and that the limit cannot be taken within linear theory, and notes that the hydrodynamic restoring force is missing from the submerged plate's equation of motion.
The mechanism: a floating plate follows the surface passively, riding the wave with minimal relative deformation. It moves with the water rather than against it. A submerged plate, fixed or constrained, experiences the wave's pressure field as a bending load against its structural resistance. The relative deformation — which drives piezoelectric generation — is larger for the constrained submerged plate than for the freely floating surface plate, even though the absolute forcing is smaller.
This inverts the obvious design choice. The energy harvest depends on the plate's deformation, not on the wave's amplitude at the plate's location. A surface plate in large waves can generate less power than a submerged plate in attenuated waves, because the surface plate is compliant (moves with the flow) while the submerged plate is resistant (bends against the flow).
The lesson generalizes beyond wave energy: in any energy harvesting system, the relevant quantity is the interaction between the harvester and its environment, not the environmental energy density alone. A stiff structure in a weak field can extract more than a compliant structure in a strong field, because extraction requires impedance mismatch. Perfect compliance means zero harvest — you're riding the wave instead of working against it.