Solar Hot Water Performance Monitor
Update: 22 Sep 14
After 3 years of operation, the sensors are failing. We are in the process of preparing our house for sale, and shortly I will de-install the Hot Water monitor.
I plan to inspect each of the temperature sensors to determine the cause of their failure. I am a little surprised at their failure, as the various installed places are reasonably protected from the weather.
I will publish the results of my findings in the near future.
A couple of years ago, our 35 year old off peak electric hot water service developed a leak in the tank and had to be replaced. I decided that going solar was a good option, particularly as after a government rebate, it would cost little more than a normal system.
So we installed a 450 litre system with three panels. After install, the cost of hot water certainly dropped a reasonable amount, but I always wondered how well the system was working, and decided to monitor the system. My initial checks on the temperature of the pipes to and from the panels showed a small temperature differential, between them, the actual temperature rising to over 100 deg C in summer. This made me wonder about the operation even more.
I decided to build a full data logging system, and measure the temperature at the top of the tank, the bottom of the tank, the feed to the panel, and the return from the panel. In addition I figured I should also measure the ambient temperature as well as the daylight intensity.
That was my starting point, and as an electronics engineer currently working in the electronics area, the design was not a problem, nor was the software. My biggest problem was finding the time, and the inclination to do it after spending eight hours a day working at the same sort of thing.
Another hurdle was that I use an Apple MAC at home, and, until I discovered the Arduino development environment working on the MAC was less than ideal. I looked at the Deumilanove and decided it was a suitable platform, but didn’t look for local suppliers, and contemplated importing some boards directly. However because of time and other distractions, I never got around to ordering them.
One day as I was googling for something else I noticed an entry for Little Bird Electronics and Arduino, so I clicked on the Little Bird link, and wow here was an Australian distributor with a great range of Arduino boards and clones, Shields,Sensors and other accessories. I was like a kid in a lolly shop.
After browsing the products at Little Bird Electronics, and comparing price and availability, I ordered a couple of Deumilanove boards, a couple of mega328 Pros, aFreetronics Power over Ethernet Shield, a handful of DS18B20 temperature sensors, a temperature/humidity sensor and a light sensor.
So far I have wired four Dallas DS18B20 temperature sensors to the Deumilanove/Freetronics Ethernet shield and mounted the board in an IP67 box on the Hot water service. I have used a weatherproof connector for the ethernet, and provided a weatherproof USB connector to enable programming without needing to open the box every time I want to update the program. The white cable coming out the bottom of the box goes to a “distribution” box where the temperature sensors and light sensor are connected.
Using the Dallas Temp – Multiple example as a starting point, I created a program to measure the temperatures. Once that was successfully running, I added the Arduino Ethernet webserver example, and now I am able to read the temperatures whenever I desire.
Ultimately, I want to use Pachube to log the performance, and to that end I will use a second Deumilanove/Ethernet Arduino to interrogate the Hot Water and transfer the information to Pachube. My reasoning for this is to enable me to add other logging points to the house, combine them and send them all to Pachube from one point.
I still need to add the software for the ambient light sensor, which I will do soon.