This is a follow up post to DIY Power Monitoring System.
Live charts of our energy usage is now available at this link . Needless to say that this project is in progress and live charts may go down at any time without notice.
Most obvious application of this system is measuring power rating of any device in the house. First, note the current rating, say 1000 Watts, then turn the device on (For example, 100 Watt light bulb) and note the increase (to 1100 Watts). To confirm the rating, do the same thing, but this time turn the device off. The total rating should go down by the same amount.
It may be necessary to confirm the rating, because there are things that go on and off automatically and it is possible that while we were measuring the power rating something else may have turned on or off at the same time.
Speaking of power ratings, some devices have fixed ratings. For example light bulbs’ power rating do not change over time. Some others like stoves, ovens, clothes dryers also have fixed ratings, but they come on and off based on their temperature sensors. Still other devices’ power consumption gradually changes over time based on their load (computers, stereos, TVs). Refrigerators produce a sharp spike when first turned on, then power level backs off and gradually goes lower.
I used a Kill-A-Watt to measure power ratings in the past. It works, but compared to this system, it is very limited. For example, how would you measure the power rating of a chandelier? The clothes dryer has a special plug and Kill-A-Watt will not fit into that plug. How about a dishwasher or an oven? These are built into the wall and there is no easy access to the plugs. Power rating of each of these devices can easily be measured with this system.
The chart above shows the refrigerator coming on. Later on when the 200 watt drop happen, the fridge was still on. The drop was because somebody turned off the bathroom halogen lights (2 X 100 watt) off.
Here is microwave turned on three different times. Notice how the data points bunch up together at higher power ratings, and longer time duration between data points at lower power ratings. You can easily see that the microwave is rated at 1500 Watts per Hour.
Here is a sample hourly chart for 100 hours. Six to eight data points lined up just below 500 Watts is when everybody is sleeping. There is a large spike early in the morning when everybody gets up. Then the rate goes back down to about 500 Watt level at about noon. The second large spike happens during dinner time and the evening. The morning and evening spikes are due to the electric stove/oven and the lighting. At the far left you can see that the rate was over 3000 Watts. That was the laundry day.
Daily chart shows our energy use at the highest level. The chart above shows 31 days worth of data. The yellow arrows mark the laundry days. Better stay out of Lisa’s way on those days. You can see that there is a slow downward trend in our energy usage.
Detecting Potential Problems
Say the freezer door to the fridge did not close properly. It would be possible to detect that problem easily because the power usage will go up from the base rate.
The base energy consumption rate is very predictable at night when everybody is sleeping. If every night the power level is at about 500 Watts, and one day it goes up to 600 Watts, you would wonder what changed to cause that increase. Hourly chart is helpful with that scenario.
This system can also be used as a parental control. I could tell exactly when my son went to bed at night the day before. When he turned the computer off, and the light off, both actions were registered in the MySQL database. No more stories about this or that.