Smart fridge (demand side management/cost reduction)

a microcontroller solution depends on the piezo crystal oscillator as its reference standard.
On an arduino nano 328 and those seem to drift by up to a minute per week,

That’s because the “328” based boards typically use a ceramic resonator vice a quartz crystal.


From the page at the link above:

Output - Crystal Oscillator provides high stability frequency output and the Ceramic resonator also provides stability output not that good as compared to Crystal Oscillator. In terms of output frequency accuracy, Crystal Oscillator provides much more accurate output than the Ceramic resonator for which parameters like temperature are a sensitive element. The accuracy for the oscillator is 10ppm-1000ppm while for the resonator is 0.1% - 1%.

Sorry I haven’t read the whole thread so apologies if it’s already been pointed out but …

Switching the power input to a fridge is not a terribly good idea. I suspect it will make the efficiency of the fridge a lot worse. Fridges have a (presumably carefully designed) cycle whereby they run the heat pump for a while and run the fan inside the fridge over a slightly later and longer period to dissipate the ‘coolth’ that has built up. Opening the door will stop the internal fan but not the heat pump for example. Cutting the power repeatedly at an arbitrary point in the heat pump cycle is probably not a good idea, I feel.

Hi Dave,

thanks for your input. I am just trying to build a concept demonstrator to engage the interest of the National Grid TSO and other parties that ought to be interested. I am definitely not suggesting that anyone else should modify their fridge at home. It needs to be built into the fridge at the design phase, to be a sensible idea.

I just think that demand side response is not as high up the agenda of the TSOs and DSOs as it ought to be. I was at a meeting of the Energy Network Association in Oxford yesterday, which mainly concentrated on flexibility services. It was 90% supply side response, and only 10% demand side. They were highlighting innovation projects involving a small number of participants being paid a few quid to switch off their heat pumps during periods of high demand. Whilst the motivation is laudable, I don’t see relying on people to manually turn-off their HPs as a credible solution to avoiding blackouts in the long term.

I think that “grid friendly” home appliances could be a low cost route to providing flexibility services, but it will require the electricity industry to talk to appliance manufacturers, and I don’t think that has happened much up to now. I may not get anywhere, but it relieves the tedium of retirement!

Thank you all for your suggestions. I now feel considerably less ignorant than a few days ago - always dangerous! I am inclined to try both, and record costs and results.

Maybe it would be safer to override the thermostat inside the fridge, that dial you turn to set the temperature. This would allow an external control circuit can tell the fridge when to be cold and when not, and provide some coarse control of the duty cycle without disrupting the compressor.


That was exactly my thought when I saw this thread. Surely, this would mean a complete replacement of the (cheap and cheerful) thermostat with something electronic and hence relatively expensive?

(I did this a long time ago as a temporary fix when the 'stat on my fridge/freezer failed - an op.amp as a comparator, transistor driver and relay, I forget what I used for the sensor, probably something like an LM355 - certainly analogue.)

I agree, overriding the thermostat would be a better idea than switching the power. But I’m not sure controlling a fridge or even a freezer’s demand is a very good idea. IIRC the main determinant of how much power a fridge uses is how often the door is open. So a red light warning people not to open the door might be as useful as controlling the electrical demand. Or control the demand of a larger heat pump - i.e. the DHW or space heating demand. That might be being designed already, of course.

It looks like the University of Lincoln and Tesco have done a bit of research into this, for retail refrigeration. There’s a paper here.

1 Like

Thanks very much for this. It’s a really good introduction to modelling a) fridges and b) grid response to demand/supply mismatches.

Interesting reading all this. My fridge-freezer consumption and when it operates has always been something I’d have liked to improve.

It does this. Runs for approximately 12 minutes every 40 minutes. Day in, and day out, doesn’t seem to change more than a minute or two on the time the compressor runs.

I did think about fitting a smart switch (shelly with power monitoring) to see if I could do a better job and reduce electricity consumption a bit.

However, I thought to myself. There’s been a group of hopefully qualified engineers who have gone through a design for this, so whatever I do will more than likely just cost me money, damage something and just be an interesting mental exercise.

And as for frequency response, by the time my fridge and a group of its friends got around to running, or not running, to control the frequency. The Fridge might not be available to RUN/NOT RUN at that exact time. The three Siemens 285MW gas turbines which are just up the road from me will more than likely beat the Fridges to it when the Gas Turbines are selected to Frequency Response instead of Load Control :slightly_smiling_face:

1 Like