Monitoring a rooftop solar hot water system

The_well_studied_HWS

David Gobbett is using a Netduino microcontroller to monitor the temperature fluctuations in his rooftop solar hot water system.

For decades, the first or only solar appliance installed by many Australian households was a rooftop solar hot water system. My parents installed one on our family home in Adelaide in the mid 1970s. In my current home we installed a conventional 300-litre rooftop system in 2006. Superficially at least, the design seemed to have changed little over the intervening years. In both cases an electric booster was connected to off-peak power, which is switched on automatically by the power meter from midnight to 7 am each day.

To reduce our energy consumption over summer, we turn off the electric booster at the main switch during late November to late March, and we still have adequate hot water most of the time. However, occasionally we unexpectedly get caught short of hot water, and at those times it’s been frustrating having no way of knowing how hot the water in the tank actually is.

Another concern with switching off the booster is that there are potential health issues when hot water system temperatures are allowed to drop below 60 °C. Lévesque et al. (2004) indicate that Legionella bacteria can grow in water temperatures up to 45 °C, but that growth stops above 55 °C, and over 60 °C the bacteria are killed. Even in hot water systems with the thermostat set to 60 °C, the lower part of the tank can remain at temperatures that are optimal for Legionella growth. It would be nice to avoid this—but that would entail having a way to sense the temperatures in the tank, which is high up on the house roof.

A project idea was sparked when a friend showed me that he was using a small microprocessor board to log solar PV power outputs. He had also connected a sensor on his water meter so he could log household water consumption. This inspired me to start on my own project to get a better understanding of what the temperatures in my solar hot water system were doing.

My interests in this project were to:
• minimise unnecessary power usage
• know when we’re running low on solar hot water, so the booster can be turned on
• minimise any risk associated with Legionella.

Setting up the temperature logging

Although I have experience as a computer programmer, I had never programmed microprocessors or worked with such things as temperature sensors. After some internet research I decided to use 1-wire devices (1-wire is a technology by which sensors and other devices can communicate). I took the plunge and purchased:
• 1-wire temperature sensors (DS18S20; 10 of these cost $18). These sensors operate over a temperature range of -55 °C to +125 °C. Several of these sensors can be connected to a single cable to form a mini network where each sensor has its own unique identification.
• a USB to 1-wire adaptor, to allow me to connect the sensors to my PC for testing (DS9490R; $28)
• a Netduino Plus microcontroller (US$70) which included a network socket and micro SD memory card slot. (See side box ‘Arduino style microcontroller boards’).

I proceeded to build the system in small steps. First I soldered three of the 1-wire sensors to a length of old telephone extension cable and then used the 1-wire to USB adaptor to connect them to my PC. Using free software (from www.maximintegrated.com) I was reassured that I had wired them correctly (phew!). Then with some extra lengths of phone extension leads, I inserted the sensors under the insulation at one end of my hot water tank and immediately saw big differences between the top, middle and bottom of the tank, as well as temperature changes in response to hot water use in the house. This was encouraging since it showed that I could get useful temperature readings from the outside of the tank.

Read the full article in ReNew 125

Prize ReNew sub banner

×