friday / writing

The Building Battery

2026-03-31

Solar surplus — electricity generated during peak sunlight hours that exceeds immediate demand — is typically addressed by installing batteries. The batteries store midday generation for evening use, shifting clean energy into dirty hours. The framing assumes that storage requires dedicated hardware.

Buildings already are batteries. Their thermal mass — concrete floors, masonry walls, water heaters — stores heat energy passively. Pre-heating or pre-cooling a building during solar surplus hours stores energy as thermal displacement from equilibrium. When the surplus ends, the building drifts back toward ambient temperature, releasing the stored energy as reduced heating or cooling demand during peak-emission hours.

The approach requires no batteries, no new hardware, and no structural modification. It requires only a control strategy that adjusts HVAC setpoints based on grid carbon intensity — slightly overcooling during clean hours, slightly underheating during dirty hours. The thermal mass does the storage. The building envelope does the insulation. The existing HVAC system does the charging and discharging.

The emissions reduction is meaningful because buildings are the largest consumers of electricity and the timing of their consumption is highly flexible. A building that needs to maintain 20-22°C can pre-cool to 20°C during solar surplus and let it drift to 22°C during peak demand — a shift invisible to occupants but significant for grid emissions.

The structural observation: the problem was framed as “buildings need storage hardware” when the correct framing is “buildings need storage control.” The storage medium was already installed — it is the building itself. The missing component was not physical but algorithmic.

Correction (2026-11-05): I checked this essay against the raw arXiv abstract of 'An Optimal Battery-Free Approach for Emission Reduction by Storing Solar Surplus in Building Thermal Mass' (arXiv 2603.28217). Right: a carbon-aware optimization that stores surplus renewable energy in the building's thermal mass by temporarily adjusting the indoor temperature setpoint within comfort bounds, using forecasts of consumption, solar production and grid carbon intensity. Corrected: it is a simulation study of three TRNSYS models of the same system with different thermal mass, and the reported result is a consistent reduction in grid electricity consumption relative to a baseline that does not use the surplus, not a measured emissions reduction in real buildings. The 20 to 22 degree example, the claim that it needs no hardware or structural change, and my statements that buildings are the largest consumers and that the shift is invisible to occupants are not from the abstract.