We live on a planet where energy is abundant.

This is a picture of an experimental solar hot water system I built in 1995, for an office I used to have before Microsoft decided they wanted to own the world of networking.  (That's another story.)  When the hot water heater failed, I decided it wasn't a problem, but rather an opportunity.  Having build an experimental hot water collector years earlier I decided I was ready to build one that had to work "for real".  This picture was of my first "production" solar hot water system. 

The basic design was that the collector box was a 4x8 insulated box with over 300 ft of black polybutelene tubing coiled inside as the absorber.  This system circulated glycol instead of water to eliminate the issue of freezing in the winter.  (Thereby eliminating the need for a drain back system.) The glycol then circulated thru a heat exchanger in an insulated 120-gallon farm dairy tank (Designed for cooling milk prior to pickup).  I had thermal sensors attached to the collector output, the tank storage temperature, the outside air temperature, and a sensor on a PV cell to read solar intensity, and the on off status of the circulation pump.  A PC on the network read the data from these sensors once a minute, 24 hours a day. 

I wrote some custom software to monitor all the incoming data and control the duty cycle of the circulation pump.  After playing with lots of elegant algorithms, simple appeared to be the best.  The circulation pump would come on anytime the collector temperature was at least 10 degrees warmer than the tank temperature, and cut off when the temperature was only 5 degrees warmer than tank temp.

On a sunny day, regardless of OAT, the circulation pump would come on in the morning and run all day.  On days of high-altitude cloudiness the pump would revert to a duty cycle depending on OAT and cloud thickness.  On days of heavy cumulus and cold OAT's it might not run at all.

For those interested in the good stuff, I have attached a sample of the data from November 4, 1995.  Raw Text file here, Excel Spreadsheet here, and more interesting a performance graph here.

The radiant loss of the tank averaged about 2 degrees per day.

The system worked very well.  I was never in need for hot water at the office.   In fact, in the summer I had to go cover 1/2 of the collector with a tarp to reduce it's energy intake when the 120 gallon tank hit 210 degrees and I was afraid it might pressurize.

This system ultimately died from being too good.  While I was on a vacation, the circulation pump failed, causing the collector  to not be able to offload it's energy.  The polybutelene tubing melted down and the system lost it's glycol.  But in the year I collected data from it I learned a lot, and can tell you today that many commercially available systems are built with incorrect design assumptions.  By the way, the ROI breakeven on this system was reached within the 1st year of it's life!

When I build my retirement home, it will feature a solar hot-water system, and, if the costs come around, solar grid-tied electric also.


As the internet will only become faster in time, the images here are in their native resolution.
You might need to enable image scaling within your internet web browser, in order to view this image properly.

All images & text 2003 Phil Case