Design

Interesting loads in the home

As promised last week, I am posting some of the results today in graphical format. The data logger has been behaving exceptionally well and all data captured was logged accurately. By knowing how each load behaves, one can make very good predictions of how much energy each one will consume and therefore design and predict accurately. In addition to that, some loads may be re-scheduled to alternative times to ensure the maximum design current rating is not exceeded for any green energy solution. Have a look at each capture below and feel free to add some comments.

f_24hrs
One can clearly see which times of the day the most energy is consumed.
f_heatpump
Notice the increase in current as the temperature of the water rises.
f_oven
Regulation the temperature is done by on-off control of the element.
f_wm-all
This is the most interesting and complicated control used of most appliances.
f_wm-wash
The washing cycle as basically running the drum for short periods at a time. Generally the ‘boring’ part of the program.
f_wm-spin
Notice the ramping up of the drum. Surprisingly, the current drawn is not very stable.

 

2 Comments

  1. Great blog, thanks for sharing. I have been meaning to build something like this for ages. I think your washing machine spin cycle current draw waveform is supposed to be a sine wave but maybe you aren’t sampling fast enough to see it. Are you going to calculate kW/h as well? Would be interesting to compare to your power bill every month.

  2. Thanks for the feedback Carl

    Yes I was also surprised by the result for those spinning cycles. The RMS calculation is done in micro-seconds, 512 readings which is averaged out on 64 samples. However, the call to logging is every 5 seconds. By then the current reading is pretty stable.
    I have also done a calculation on the kWh but still need to apply a correlation factor. It gives out a pulse on the box so we know when how and when a major load is active.

    Jacob

Leave a Reply

Your email address will not be published.