Black Plastic Pipe Solar Hot Water
This project was a quick and dirty experiment to see how much usable heat could be generated by placing a coil of black plastic (polyethylene?) pipe on a roof. It also served as a shakedown for the TS7260 hardware and software. Since the collector had to be located above the hot water tank, a circulator was required. The TS7260 monitors temperatures and controls the circulator pump.
The system consists of the following components:
The 'collector pipe sensor' measures temperature near the top of the collector. If that temperature is more than 3° C warmer than the tank bottom temperature, the circulator pump is turned on. The circulator pump draws water from the bottom of the tank and circulates through the PVC coil, where it then returns to the top of the tank. The pump flow is limited by a valve which restricts the flow rate to a relatively low (but unknown) value so that water passes through the coil slowly.
The TS7260 measures temperatures once per second. A control process runs every 35 seconds and determines whether the pump should be on or off.
A separate process logs data to a database on a remote computer. A snapshot of all inputs and outputs is logged every 30 seconds. This data can be viewed here.
The experiment was a success in that it routinely provided water hot enough for showers, and dramatically reduced the load on the on-demand heater. If the controller is already available (for wood heat control, for instance) then the additional cost to add solar hot water using this approach is very low - less than $150.
The project also generated it's share of issues and learning opportunities. For instance, the circulator pump does not provide enough flow restriction to prevent reverse thermosiphoning. After the sun sets, hot water from the tank rises, flowing backwards to the coil on the roof, where it cools, flowing through the pump and back to the bottom of the tank. A check valve has been installed to prevent this.
At first, the control task ran every two seconds. Despite logic designed to avoid the problem, there were times when the pump would cycle too frequently. Since datalogging happens at 30 second intervals, this cycling is not readily apparent when examing the data. The control task was modified to run every 35 seconds, so that every pump cycle is captured. 35 seconds also provides enough time to move a reasonable amount of water through the collector.
Thermal stratification in the storage tank is critical to efficient operation. Best results are obtained if the water in the bottom of the tank is cold and the water in the top of the tank is hot (and therefore useful for showers and such). The inlet at the top of the tank is positioned such that incoming hot water is injected straight downward, causing some mixing and temerature inversions. As of September first, the inlet has been moved to a better location with a baffle to minimize mixing.
A review of the data shows that the 30 gallon tank does not provide enough heat storage capacity. It is often near the maximum temperature that it can attain by early afternoon, so potential heat is lost. Maximum hot water usage happens after sunset, and the tank is depleted before the demand has been satisfied. Planning to use hot water in the late morning or early afternoon can help, but a larger tank would be ideal.
Finally, the tank is not well insulated. Perhaps because it is designed to be heated with propane and therefore has a flue running through it, the tank loses more than 10°F during the night. This takes quite usable water at 120°F and makes it marginal at less than 110°F by morning.