Solar Water Heater

Three elements: tanks, a glass-covered box, and pipes carrying water from one tank to another. Test them one at a time.

Use three identical containers, each with 200 ml at the same starting temperature. Record that temperature.

Paint the outside of one container matt black. Leave the second bare and the third bare — the third is for the box test.

Same place, same sun, same 30 minutes. Measure both. Black absorbs more. Record the gap.

Make a plywood tray a little larger than the bottles and line the inside matt black. This is Kemp's box "secured on the roof".

Lay a float glass sheet over the box as the "glazed cover". Glass lets sunlight in and holds warmed air in.

Same sun, same 30 minutes, black bottle now under glass. Measure. Compare against step 4 — the glazing is worth a large share of the gain.

Take the lid off and keep measuring every five minutes. The rise slows immediately — warm air is carrying the heat straight off.

Join three containers with copper pipe so water must flow through all three in turn. This is the claim that makes it a heater rather than a hot tank.

Run water slowly through and measure at each junction. The temperature climbs in steps — each tank adds to what the last one did.

Run twice as fast. The outlet is cooler. Slow flow, hot water; fast flow, warm water. The collector delivers power, not temperature.

Turn a tank so its widest side faces the sun, then edge-on, measuring each. Kemp made his tanks elliptical and laid them with the wide axis horizontal "to increase the surface exposed to the sun".

Measure again an hour after sundown. It has gone cold. The Climax had no insulated store, and that single fact shaped the whole later industry.

The patent. US 451,384, "Apparatus for utilizing the sun's rays for heating water", granted 28 April 1891 to Clarence M. Kemp of Baltimore, Maryland. Kemp sold it as the Climax, and it is generally regarded as the first commercially marketed solar water heater. He states the purpose without any grandeur: "to utilize the sun's rays to heat water for bathing and other domestic purposes".

The drawings are worth reading as a system, not a device. Figure 1 shows the box on a house roof, piped down to a bath-tub, with a cock controlling flow into the first tank. This is not a laboratory demonstration of solar heating — it is a plumbing installation, drawn into an actual house, with the tap at the end. That is what separates Kemp from a century of earlier hot-box experiments: he treated sunshine as a utility connection.

Three design decisions, all still in use. First, glazing: the tanks sit "within a glass-covered box", which admits sunlight and suppresses the convection that otherwise strips heat off a hot surface — step 8 shows how much that is worth. Second, series flow: pipes carry water "from one tank to another" so each tank raises the temperature further, which is why the outlet is much hotter than any single tank would be. Third, shape: the tanks are "preferably elliptical in cross-section" laid with the larger diameter horizontal, purely "to increase the surface exposed to the sun". Kemp is optimising collector area per unit of stored water, which is precisely what a modern flat-plate panel does.

The honest weakness is the one step 13 finds. The Climax stores its water in the collector, so the store is also the thing radiating to the night sky. It heats beautifully by day and is cold by morning — fine in southern California, where it sold well, and much less useful in a cold climate or for an early bath. The fix was to separate collecting from storing: put an insulated tank indoors and circulate water through an outdoor collector, which is what the later Day and Night system and every modern installation do. Kemp's patent contains the collector but not the store, and knowing which half is missing is the useful part of reading it.

Why it matters now. Domestic hot water is a large share of household energy, and it is a low-temperature demand — which is exactly what a cheap collector supplies well. Solar water heating remains, per unit cost, one of the most effective solar technologies there is, and the arrangement in this 1891 drawing is recognisably the same machine.