The Cambridge Clarion’s solar-kettle page is a good antidote to the modern Internet’s endless supply of projects that somehow work perfectly on the first try. This one boils water, measures what happened, admits what went wrong and then spends most of the page investigating the failures.
Visit the home-made solar kettle article.
Stephen Hewitt published the experiment in October 2022. The device is a one-meter-long parabolic trough about 80 centimeters wide. A reflective sheet concentrates sunlight onto a 22 mm copper pipe containing roughly a mug’s worth of water.
Under good conditions it could heat that water in around ten minutes.
That is the headline result. The much more interesting part is everything after it.
The page measures instead of merely claiming
Temperature was recorded with a K-type thermocouple inserted into the pipe. The article includes heating curves, a cooling curve and estimates of net heating power and heat loss.
Around 50°C, Hewitt estimated at least roughly 215 watts per meter of net heating power under one test condition. He then worked through the implications of heat loss near 90°C and estimated the optical efficiency of the reflector, while repeatedly noting the assumptions and uncertainties in those calculations.
This is old-school web publishing in spirit even though the article itself is much newer than GeoCities. The evidence is on the page. You can see the apparatus, see the graphs and follow the reasoning.
The kettle also tries to attack the operator
The prototype has problems.
The most dramatic is that it can eject much of its water when boiling. The page plainly labels that as hazardous and says the design in its photographed form is not suitable for taking water all the way to boiling. The short spout can also dribble scalding water as the contents expand.
Other problems are less theatrical but more instructive: water temperature varies along the copper tube, some reflected sunlight misses the pipe, the reflector geometry is imperfect, filling and pouring are awkward, and thermal expansion of the polypropylene reflector may alter its shape.
Each flaw becomes another engineering question instead of being cropped out of the photographs.
Why this belongs in the Ancient Web pile
The best independent technical pages are not valuable because they are old. They are valuable because they preserve the full thought process that commercial content usually compresses into a list of tips.
This page records dimensions, materials, measured curves, wrong assumptions, physical hazards and proposed revisions. It is exactly the sort of document another experimenter could discover years later and use as a starting point rather than repeating every mistake.
That was one of the Web’s original superpowers: strangers leaving useful laboratory notebooks in public.
Read the Cambridge Clarion solar-kettle experiment and enjoy a project honest enough to document the part where the kettle spits boiling water back out.