History of the Williams Farm All Solar Powered Hot Water System
graph button
LIVE System Performance Graphs
See our real-time data updated once a minute
120 - 50 = DELTA T 70F
Click HERE
for Solar Hot Water Math Calculations
Iwalani's Voyage Around the World
Weekly logs of Iwalani's three year circumnavigation written by Philip Shelton, Amy P. Wood and Stewart the Cat.

small wv book

The "World Voyagers" Book
A true story of the three year circumnavigation by Philip Shelton, Amy Wood and Stewart the cat. From designing and building a 42 foot wooden cutter "Iwalani" to sailing around the world— this is not a watered down, sugar coated tale, but a "no holds barred" account of just what it's like to live a "dream."

After returning from our circumnavigation in “Iwalani” in 2003, we were determined to cut back on our use of non-renewable energy. Living on a boat for three years taught us the value of resources. We had a Wind Baron wind generator on Iwalani that was rated for 600 watts. I say, “rated” because we were under the misconception that the generator would be putting out 600watts. After living with it for three years, we learned first hand just what a wind generator will and will not do.

Wind generator output, as told by sales people, is the maximum output with a wind speed of 25mph. Because the output is logarithmic, the output at 12 mph is about one fourth or 120watts, not one half or 300. On top of that, wind is not steady. So to average a wind speed of 25mph, you need winds ranging from 20-30mpg without ever dropping below 20mph. That just doesn’t happen.  With wind speeds below 8mph, the generator doesn’t put out any electricity at all. None.

Knowing all this, we were still hopeful that we would be able to produce some electricity at Williams Farm with wind. To that end my son, Nathaniel and I designed and erected a 65ft tower made from 2” EMT tubing. We attached a wireless wind speed recorder and began gathering data. Much to our dismay, but not surprisingly, we discovered that the average year-round wind speed (recorded over a three year period) here in Georgetown, 5 miles from the open water of the Gulf of Maine, was a paltry 5mph. That ended the wind power idea.

Amy thought that perhaps using the sun would be better. Even though we don’t have sunny weather like Arizona, we decided to look into it. I bought a solar irridiance data logger and studied the output. It soon became clear that the sun was more reliable than the wind, even though it is available only half the time. Solar power doesn’t work at night. Go figure.

Photovoltaic panels were still as expensive as ever and grid-tie regulations were yet to be born, so we decided on solar hot water. The original push was to use solar hot water, connected to a radiant heat system, to offset the LP gas heater in our greenhouse that was costing us hundreds of dollars during January and February.

We visited a local “green” expo and spoke with “experts” in solar hot water systems. I looked their systems over and saw lots of AC pumps and controls. It reminded me of the guts of the space shuttle. I believe in the KISS (keep it simple stupid) design philosophy. I thought that a solar powered pump would work in harmony with the output of the collector. Sun comes up, collector heats up, the pump comes on and moves the heat generated by the collector to the storage tank. The sun goes down, the collector cools off and the pump stops. How much simpler could it get?

I was told, in no uncertain terms, that solar powered pumps would not work. They don’t start up soon enough in the morning to keep the collector from over heating.

Hummm I thought. “Can’t be done.” Those words are like gasoline on a fire. My brain goes into overdrive and doesn’t stop until I “make it work.”

I made a few decisions, we would use evacuated tube collectors to eliminate energy loss during cold weather, we would have a large storage tank to hold over some of the energy during cloudy stretches and the entire system would be powered by the sun. No using fossil fueled, power plant electricity in our attempt to be “green”.

I decided to start with a single 20 tube collector, figuring that I could always add more. I was skeptical of the claimed output, after our experience with wind generators, so wasn’t ready to make a large investment. I put the collector at a tilt of 67 degrees. We were looking to get the maximum out of the system in the winter months, so a tilt for the winter sun angle made the most sense. Another benefit was shedding snow and keeping the summer output to a minimum so the system wouldn’t over heat as much.

The tank size was determined by the space we had in the basement. A 328 gallon tank would just fit between the chimney and the foundation wall. Not exactly scientific, but when you don’t know anything, you don’t know if your making a mistake.

A closed loop system seemed like a good idea. You don’t have to worry about getting the piping installed so that all the HTF (heat transfer fluid) drains back (severe limits on pipe run design) and you don’t need over sized pumps to get the system started up. Drain back pumps have to be able to push the HTF to the highest point in the system against gravity. Closed loop systems do not have to fight gravity, so can use low powered pumps. The HTF in our case was 50% polypropylene glycol (non-toxic antifreeze) and 50% water. This was also used in the radiant heat loop, so the greenhouse could be allowed to freeze without damaging the heating system.

I wanted to be able to log data from the system, to know just how well it was working. Sure, I could have done the math and come up with some numbers, but I just don’t trust theory. I wanted proof.
Data loggers can be very expensive. Thousands of dollars expensive. I spent some time surfing the Internet and came across the K145 serial data logger for temperature. Using the serial port of a computer, a build it yourself electronics kit and a simple interface like hyper terminal, you can log temperatures. The kit was $30 and the sensors are $5 each. So, for $50 I could log 4 temperatures. I eventually bought 4 kits and monitored 16 temperatures. I soon tired of producing excel spreadsheets and charts with the hyper terminal files and used free software from John Gray at to make the logs easier to work with.

There were some minor issues with the kit and software. Apparently I was stretching the limits of applications for the K145 with sensor runs of over 100 ft and long term (months at a time) logging. But, things worked out and information began to pile up.

I started the system up in March of 2009 and soon discovered that the highest tank temperature I could produce was 90 degrees. Not really enough for the radiant heat system, but high enough to add some preheat to our domestic hot water supply.

This was something I hadn’t even thought about. Well water comes in at 50 degrees and needs to be heated to at least 120 for domestic hot water use. By having the well water run through a heat exchanger in the solar storage tank, we increased the temperature at least 40%. That meant at least a 40% savings on our domestic hot water. In our case, home heating oil for our tankless hot water heater.

The data logging showed the BTU output of the collector was right in line with the manufactures claims. I did the analysis when there was no household hot water use. One twenty-tube collector raised the temperature of 328 gallons of water 10 degrees on a sunny day. A BTU equals a one degree Fahrenheit increase in the temperature of 1lb water. 328 x 8 (a gallon of water equals ~8lbs) = 2,624 lbs of water 2,624x10 = 26,240 BTU’s. Click HERE for more solar hot water math.

So with this encouraging information, we installed two more twenty-tube collectors. Now we had a total of sixty tubes.

A direct connection of the 40w solar pv panel to the 20w pump was a problem with the new 60 tube array. (You have to double the pv wattage, so the pump has more power to move the polypropylene glycol at the same rate as water.)

PV panels are peculiar in that any shading of the panel, even the size of a small leaf, reduces the output dramatically. Also, they don’t all put out the same power at the same rate, even though they may be equal in wattage output. Some of this is due to the materials used and also how the manufacturer manipulates the output.

Knowing this and the fact that the early morning and late day sun was hitting the pv panel and the collectors from the back side (that’s right, the sun rises north of east and sets north of west in the summer), I decided to experiment. This involved coordinating the output of three 10w collectors connected to a 10w pump.  Why the 10w pump? Turns out that the 10w pump put out the correct flow for a 60 tube array, according to the collector manufacturer. Why Silicon Solar recommended a 20w pump for a 20 collector (my original order) I don’t know.

I had my son Nathaniel build a square box with three adjustable sides so that we could angle the three panels from 0-45 degrees up and down and swing the two opposing sides from east to south and west to south. Sounds a bit confusing, but what I was trying to achieve was solar power early in the morning and late in the afternoon. Because no more than two panels would be facing the sun at any time, I wouldn’t be over powering the pump with all 30watts.

This system worked better than my expectations. Now I had a chance to change the angles of the pv panels to vary the output and see what would work best. While the system worked OK, I wanted to have even more control over just when the pump came on and off. The only way to do that was to use a differential controller. IMC makes one that works on pv power, so that’s what I bought.

So now I had a multifaceted pv array and a solar powered differential controller. The controller kept track of the collector temperature and the tank temperature. Once the collector reached 10 degrees warmer than the tank, it turned the pump on. My tank temperatures were rising 25 degrees a day and it soon reached 120 to 140 degrees every day. Now we had plenty of hot water. As a matter of fact, we turned off the furnace for the remainder of the summer and didn’t turn it back on till the middle of November.

It’s a good thing I was collecting data because I discovered that having the DHW (domestic hot water) heated by the solar storage tank and then running through the now cold tankless hot water heater, the output was cooled significantly. It was time for more plumbing work. We put in a three-way valve to by-pass the tankless heater when it was shut off. Problem solved.

When we went on vacation in August we had the tank rise to over 160 degrees. To prevent the tank from over heating, I bought another electronics kit called the K190 relay controller. I set it up so that when the tank reached 160, the radiant heat pump would come on and dump the heat into the greenhouse floor. Worked like a charm!

I also used the K190 to control the radiant heat loop in cold weather. It’s set up so that the relay for the tank temp and the circulator pump are in series. When the tank temp drops below 70, the relay opens and prevents the radiant heat loop from dropping the storage tank temp to 50 degrees. It’s a decision I made to keep some of the storage tank energy for the DHW.

The radiant circulator pump is a 12v el-sid, run off a deep cycle battery that is charged by the original 40w pv panel.

At first I had the radiant pump relay temperature sensor in the sand half way between pex loops, but the sand heated up so fast that there were wide swings in temperatures and the circulator pump was cycling on and off every 15 minutes. I pulled the sensor out of the sand and exposed it to the air a few inches above the floor but that didn't work very well either.

It takes the floor about four hours to come up to a stable temperature, so the floor heat was four hours late in offsetting the drop in the greenhouse temperature. To better anticipate the temperature drop, I placed a sensor outdoors connected to a seperate relay. Now the pump comes on long before the greenhouse cools off.

I now have the inside temperature sensor relay in parallel with the outside temperature sensor relay. That way, on cloudy days when the outside temperature keeps its relay open but the temperature drops inside the greenhouse, the radiant pump still comes on.

Because I wanted to generate web graphs from my data, I pursued sending my data to our web server and generating graphs. Turned out to be a bit more complicated than I thought. I started looking for alternative solutions. There were services and hardware you could purchase, but they were out of my budget. I kept looking and soon discovered the WEL web energy logger. It had DIY option and no monthly subscription. And I could use the sensors I all ready had.

I soon had it set up and generating web graphs. It also displays live data onto a diagram of your system. You can monitor motor on/off cycles, electrical usage, pulse cycles, sun values, humidity, accumulate data by day/month and year and combine data into simple formulas. Now I could really generate some useful information! Check out our live data graphs HERE.

At present we are heading into our first winter season for the system. I’ll be posting updates on the home page on how things are going.

Mouse Over the Image Above for a LARGER View of the SkyCam Picture
Updated Every 15 Minutes


Check the progress of Phil's Tundra Kit Plane


Paintings, Artwork and Blog of Dr. Amy Peters Wood.


Building the 65' schooner "Janet May"