Offgrid Fridges

An Eccentric Anomaly: Ed Davies's Blog

The other day I was asked what I'd be doing about running a fridge off-grid. This was once an interesting question but it's not so much now though the explanation of why is sufficiently involved that I felt the others present might have been less than completely enthralled so I just said “plug it in”. Here's the longer answer.

Back in the 1970s or even the 1990s people went to some extremes to do off-grid refrigeration. E.g., they'd wrap the fridge in extra layers of insulation or, in colder climates, freeze water, salt water or oil outside and bring it in to keep the fridge cold without using power.

Now that sort of thing is probably not the best use of one's money and effort. This is mostly because PV panels have got a lot cheaper and standard fridges have got a lot more efficient.

Here's my domestic electricity consumption last night from just after I turned off the bedroom light until just before the central heating came on. It links to a readable version:

The base level of power consumption is about 16 watts from the various bits of “always on” kit - ironically mostly on to log the power consumption (and also for telephones).

The big increases to 70 watts or so for about 20 minutes are, of course, the fridge running. Notice how they get a little further apart as the kitchen cools overnight. The first pair are only a few minutes over an hour apart increasing to about an hour and twenty minutes for the last pair; the fridge came on again just after the boiler fired up at 07:00 but if I showed you that the Y axis would be a bit more squished.

(The spike at 05:30 is lights for me to visit the loo. Yeah, this logging quickly moves into TMI territory.)

So, the fridge consumes about 60 watts for about 20 minutes in each 1:20 period for an average of about 15 watts. During the day it'll consume a bit more as the door is opened and closed, etc, and the kitchen's a bit warmer.

This is just a standard tall fridge/freezer supplied by my landlord. It's a Linsar FF800 which is A+ rated (i.e., basic; Argos only do A+ plus a few A++ and A+++; thank you whatever mega brain set up a scale starting at A for the most efficient when the goal was to increase efficiency) with a notional consumption of 231 kWh/annum (26 W) which is reasonably consistent with the above.

Roughly speaking, the capital costs of winter-time off-grid electricity is £30/watt (unless you have a stream you can put a turbine in when it can be noticeably cheaper). PV costs about £1/W (quite a bit less for the actual panels now but you need fixings, cable, charge controllers, etc) and generates for an equivalent of about 45 minutes a day on average so about 1/30th of the time.

So, to eliminate the load presented by this fridge would be worth about £30/W × 15 W = £450. Definitely not a reason for spending upwards of a thousand pounds on a high-spec fridge which might only halve the consumption anyway.

In general, my approach to thinking about pretty much any energy saving scheme is to first try to work out the relative cost of just adding more PV.

This is where things get involved, though. Suppose, for example, that you have a choice between two fridges:

(B) costs £200 less for the same performance so you pick that, right? Hang on a minute:

The extra 10 watts of heating is costing us £200 so £20/W. This is cheaper than the assumed £30/W for PV because we bought a cheaper fridge. I've previously shown that under adverse conditions (low insolation, high output temperature) PV can beat solar thermal panels. While space heating doesn't require the high output temperatures applicable there this effectively cheaper PV swings the equation slightly towards PV.

Incidentally, this space-heating effect is also relevant if you're living in a wooden cabin in Montana and reckon the way to run your fridge is to freeze some plastic milk churns of water outside then bring them in to melt in your fridge. Every cycle you're bringing coolth into the house (dumping heat out). If you're in a bitterly cold climate and a poorly insulated house this probably doesn't increase your heating requirements much (you're burning a tree a day, anyway) but in a well insulated house in a temperate climate it could be quite significant. I don't know how much heat flows in and out of a fridge but it could well be quite a bit more than the electricity consumption needed to pump it out.

Something that's a bit of cheat here is that I've ignored the batteries which have a limited life and are not cheap. What it might be useful to do is to somehow cool the fridge more when power is available (the sun is shining or the wind is blowing) and store coolth so as to avoid or delay running the fridge from the batteries as much as possible.

However, even quite a small wind turbine will cover the fridge and other base loads many nights and batteries should be getting cheaper so this is likely not something to worry about too much.

Update: 2018-11-01: SteamyTea emailed some measurements he did of his fridge earlier in the year. This was a fairly cheap under-counter fridge in a chilly overnight kitchen (around 11 to 12 °C). It ran about 6 minutes in each hour and took, when running, 49 W so the average consumption was only 4.9 watts, which really isn't a lot, though this is probably about as low as it gets for standard fridges.