Would repulsive gravity necessarily violate energy conservation in ways classical mechanics couldn't explain?

This is a subtle question that hinges on what we mean by "energy conservation" in gravitational systems.

In classical mechanics, energy conservation flows directly from time-translation symmetry: if the laws of physics are the same yesterday as today, energy is conserved [AG-2025.12-638]. Repulsive gravity in that framework would not by itself violate energy conservation—a repelled object gains kinetic energy, which must come from somewhere (e.g., a potential energy field), just as it does for attraction.

However, the picture becomes more interesting in general relativity and beyond:

In general relativity itself, energy conservation is already subtle. In curved spacetime, strict global energy conservation fails because spacetime lacks universal time-translation symmetry [AG-2025.12-638]. This is not a flaw—it's a feature. Gravitational potential energy can be converted to other forms without a global "energy account," and this is well understood. Repulsive gravity within GR would fit into this framework without introducing new violations [AG-2024.02-457].

In modified theories, repulsive gravity can coexist with explicit energy-conservation violations. Horava-Lifshitz gravity, for instance, allows time-varying Lagrangian parameters that break time-translation symmetry, leading to stress-energy conservation violations at high energies [AG-2024.08-134]. But these violations exist *independently* of whether gravity is repulsive or attractive—they're a property of the theory itself.

In quantum gravity regimes, repulsion can emerge from quantum interference effects without violating energy conservation in the traditional sense [AG-2026.02-268, AG-2026.02-534]. The repulsion arises from superposed gravitational fields, and energy is still accounted for; the quantum nature of spacetime itself allows novel force configurations.

The key insight: Repulsive gravity does not *necessitate* energy-conservation violations. Classical mechanics can accommodate it. General relativity already permits repulsion (velocity-dependent terms in certain geometries) [AG-2024.02-457] without new violations. Violations, when they occur, come from *other* features of the theory—broken symmetries, quantum effects—not from repulsion per se.