concepts · updated 2026-07-19

Thermosiphon

confidence: high volatility: cold

A thermosiphon is a passive circulation loop for water or other fluids driven by temperature-induced density differences — no pump, no moving parts. Warmer water is less dense and rises; cooler water

A thermosiphon is a passive circulation loop for water or other fluids driven by temperature-induced density differences — no pump, no moving parts. Warmer water is less dense and rises; cooler water sinks and displaces it. With the right geometry, the loop sustains itself as long as a temperature differential exists.

In Baer’s archive the thermosiphon is not incidental — it is the mechanism that makes “no electricity for cooling” possible. Kayatekin’s 2025 paper identifies the double-play thermosiphon as Baer’s central contribution, seeded in the late 1960s and still evolving at his death in May 2024.

How it works

Density-driven circulation requires two things: a temperature differential and a closed loop oriented so that the warmer fluid can rise and the cooler fluid can fall.

         [roof radiator / absorber]
               ↑ hot rises
               |           ← cooled water returns down
         [storage tank / ceiling reservoir]

For cooling (the dominant use in the archive): the roof radiator is physically higher than the interior storage. At night, radiation to the sky cools the roof panel. Cooler water is denser, sinks, and displaces warmer water from the tank upward to the roof. The loop runs continuously without a pump until temperatures equalize.

For heating (south-wall variant): an absorber is lower than the ceiling tank. Solar gain warms the absorber, water rises to the tank, cooler water returns down. Because the absorber cannot drain back (it sits below the tank), it must be freeze-tolerant.

Self-regulation: During summer days the roof surface is too hot from direct sun for thermosiphon to begin — it would move hot water into the building. Circulation simply stops. No valve, no control, no electricity. This is a feature, not a limitation.

Applications in the archive

Cool Cell (c. 1999–2002)

The Cool Cell’s two primary modes run by thermosiphon:

  • Summer night cooling: Roof radiator/absorbers cool by night-sky radiation. Thermosiphon circulates the cooled water to the ceiling reservoir, recharging it with “coolth” overnight. Winter solar heating uses a small low-power pump instead — thermosiphon is not used in this mode because the roof is the heat source and the direction of flow is reversed.
  • Winter south-wall heating: In colder climates, a freeze-tolerant south-wall absorber heats by thermosiphon with no pump. The absorber sits below the ceiling reservoir so cold water drains down naturally.

The Cool Cell brochure describes the summer-day state as: “Roof-mounted radiator/absorbers are too hot from the sun for thermosiphon circulation — no solar gain enters the system.” Self-regulation at no cost.

Shower test (2003)

Baer’s field test, described in the NMSEA SunPaper: a 30-gallon plastic drum above the shower, gravity-fed, with thermosiphon lines running down to unglazed rooftop collectors. Sky Mat polypropylene panels produced warm showers through icy January weather in Albuquerque (high 46°F). This is a domestic-scale, pump-free implementation using only height difference and density.

Bruce Davis studio (2004)

12 polypropylene Sky Mat radiator/absorbers (127 sq ft) on a garage roof cooled a 236 sq ft studio through 95°F summer days. Circulation was by thermosiphon — no pump. Storage: 42 vertical 4″ PVC drain pipes holding 200 gallons with 300 sq ft of surface area. Davis: “great satisfaction: no noise, no fan and no added moisture.”

This is the most thoroughly documented thermosiphon installation in the archive.

Drop City air loop (c. 1967–68) — earliest known system

Baer’s on-camera claim in the Drop City documentary (c. 2002):

“It took advantage of the buoyancy of warmed air to power the cycle that took the warmth through the rocks, store the heat in the rocks, and then curl could open a little door. It was the first convective air loop rock storage system that worked, that I know of. This is still a darn good way to use the sun. Nothing wears out.”

The system: south-facing slope of a dome, solar collectors made from salvaged car rear windows, river rocks as thermal mass, natural air convection driving the loop. No water — an air thermosiphon. A resident describes the dome as “perfect for solar panels” and says “Steve’s idea was to get river rocks.” This predates the Baer House (1971) and the water-based systems by several years, and is the archive’s earliest documented Baer solar installation.

Nick Pine’s 1995 Usenet post quotes Sunspots (1979 Cloudburst ed.) on air-loop rock storage systems — the Sunspots chapter likely describes the same principle formalized from the Drop City experiment.

1973 bubble engines

The October 1973 issues of the Tribal Messenger describe low-temperature-difference engines using immiscible liquid pairs and buoyant bubbles — an early thermosiphon-adjacent idea. The same density-driven logic appears here, formalized into a heat engine rather than a building system.

The exception: pool heaters

Pool-heater cooling is the archive’s clearest example of thermosiphon’s limits. Pool collectors present too much flow resistance for passive circulation to overcome. A small pump and controls are required. The comparison with the Cool Cell (which needs no pump for cooling) is explicit in the source: “the pool collector has too much flow resistance for thermosiphon.”

Flow resistance is the practical constraint. Short, large-diameter water paths with modest height differences are thermosiphon-compatible; long runs with many fittings or small-bore tubing are not.

Why Baer kept returning to it

The archive returns to thermosiphon across five decades for the same reasons:

  1. No electricity. Nothing to power, nothing to fail.
  2. No moving parts. Long service life; no maintenance interval.
  3. Silent. “No noise, no fan” — Davis.
  4. Self-regulating. Circulation stops when it would be counterproductive (summer day).
  5. Geometry is the design. The system’s behavior is determined by heights and pipe sizes, not by programming or controls.

This aligns with the archive’s broader argument: passive systems are more durable than active ones because durability is structural, not mechanical.

See also