How does the thermodynamic cost of maintaining non-equilibrium compare to the extractable Casimir force output?

# Thermodynamic Cost vs. Casimir Force Output

This is a sharp question that cuts to the heart of whether Casimir-based propulsion systems could ever be practical. Let me map what the archive tells us.

The Cost Side: Maintaining Non-Equilibrium

The thermodynamic price tag comes directly from Landauer's principle. When you continuously monitor and record quantum states—which is necessary to extract work from vacuum fluctuations—you must dissipate heat proportional to the entropy of your measurement record [AG-2025.12-670]. In other words, every bit of information you log costs you heat, and you cannot avoid this by going slower or cleverer; it's a fundamental lower bound.

More broadly, any quantum measurement apparatus that creates system-apparatus correlations, collapses superpositions into definite outcomes, and resets itself requires minimum work input tied to information-theoretic quantities like measurement efficiency and completeness [AG-2024.02-2018]. This applies whether you're measuring photons in the vacuum or monitoring which electromagnetic modes are excited.

The Output Side: Casimir Forces from Non-Equilibrium

The exciting part is that thermal non-equilibrium *does* generate harvestable Casimir forces. When a biased semiconductor is kept out of thermal equilibrium with its surroundings, it emits radiation that exerts a *repulsive* Casimir force on nearby surfaces—forces strong enough to be detectable with current experimental apparatus at submicron scales [AG-2024.03-1773]. Similarly, an inhomogeneous body held at a different temperature than its environment experiences spontaneous self-propulsive forces and torques without requiring exotic materials [AG-2024.05-941].

The physical picture is that nonequilibrium breaks the symmetry between attractive (virtual) and repulsive (real) vacuum contributions [AG-2025.01-1012].

The Unfavorable Comparison

Here's the catch: the archive does not directly compare the *rate* at which you must dissipate heat to maintain the non-equilibrium state against the *magnitude* of the extractable Casimir force output. Critically, maintaining a biased semiconductor or a temperature gradient itself has a thermodynamic cost—one that likely scales unfavorably with the smallness of the resulting forces. The Casimir forces discussed are typically piconewton to nanonewton scale [AG-2024.03-1773], and achieving the required non-equilibrium in confined geometries would demand continuous power input. Unless that input power is orders of magnitude smaller than the work equivalent of the extracted force (an efficiency question the archive does not address), the system would be thermodynamically uncompetitive.

A rough analogy: running a heat engine to pump current through a biased semiconductor to generate a repulsive force a nanometer away sounds like it has a very poor coefficient of performance.

The preprints establish that both costs and benefits exist, but do not quantify their ratio for any concrete system.