How Much Solar Power Do I Need To Run A Refrigerator?
Some weeks ago I posted an article examining the possibility of running an ordinary household fridge with solar panels and concluded, like most other solar bloggers, that it generally needs about 4 solar panels for the average refrigerator.
Checking back, I’m finding that many treatments on the internet often deal in theory and use average values with ‘ruleofthumb’ calculations for power, running currents and duty cycles.
In this post I take two refrigerators, one old and one new of a similar size, and measure actual current values over 24 hours before sizing the solar panel array needed.
Page Menu
How Many Solar Panels To Run A Refrigerator?
On average two solar panels rated at 100 watts will run it for most of the summer. Most locations will need four 100 watt solar panels to operate the average fridge during winter.
The test results show that even in my very sunny location in July a 100 watt solar panel is not going to run a refrigerator. The exact number and ratings can be extrapolated using the calculator and solar irradiance sites in this post.
This post contains affiliate links to Renogy solar products. Solar Empower only partners with the best solar suppliers based on product tests.
The Refrigerators Used For The Solar Project
I have two similar sized refrigerators in different parts of the house. The 10 yearold Whirlpool model sounds a bit clunky but works OK, while the newish AEG is smooth as silk.
For this project I’ll compare the two and see how much energy they draw over a period of time before working out how many solar panels we would need to run each one.
How many watts does a fridge use?
Conventional wisdom tells us that a modern fridge is rated between 100 and 400 watts, but I can’t find anything on the nameplates to verify this or not.
In any case, I’m far happier taking measurements and using actual readings from working appliances.
The instrument used is a simple AC wattmeter showing volts, amps, watts, watthours, frequency and power factor.
I’m really only interested in current, watts and watthours, but the power factor reading for the older fridge surprised me.
The new AEG refrigerator uses 78 watts while the old Whirlpool (tested below) uses 154 watts.
Before moving on to sizing solar panels for running refrigerator, we need to talk about the electrical characteristics of appliances like fridges, freezers and air conditioners that contain compressors with motors.
First of all, anything with a motor has two possible current draws: steady state running and startup, or surge current. The ratio of the compressor’s running time to the ‘stopped’ time is called the duty cycle.
Typically, the duty cycle is taken as 50% or 0.5, that to say the motor runs half the time.
The actual duty cycle value depends on ambient external temperature, internal temperature setting, how full it is and how often the door is opened for access.
It’s generally less than 50%, but this figure errs on the pessimistic side, so we don’t end up too tight with the solar panel sizing.
What is power factor and why is it important?
An unfortunate characteristic of a.c. circuits is a phenomenon known as the circuit power factor.
This is expressed as a decimal number less than unity and describes the phase relationship between voltage and current.
In an ordinary resistive circuit, such as an electric heater or hot water cylinder, the current is in phase with the voltage. They are said to be ‘in phase’.
 In this case volts (V) x current (I) = Watts (W)
However, whenever an a.c. circuit contains loads with a coil, like a motor, magnetic circuits are created with which have the effect of pushing back the phase angle of the current waveform. The current is said to ‘lag’ the voltage.
In this case, volts x amps does not equal watts but the Power Factor value is used as an extra multiplier. Because the peaks of voltage and current are not at maximum at the same time, then more current is needed for the same watts.
Let’s say the PF of a circuit is 0.8 lagging (means it’s inductive) and it draws 8 amps at 230 volts (for Europe):
 volts x amps x PF = Watts
 230 x 8 x 0.8 = 1472 watts
The motor is rated at 1472 watts but is drawing more current than it normally would due to the poor power factor. If the PF was unity the equation would change. The motor still needs 1472 watts but it will draw less current:
 watts/volts = amps
 1472/230 = 6.4 amps
If this circuit’s PF was closer to unity then less battery capacity and solar power would be needed to run it.
How to improve power factor
Fortunately, there is a way of improving power factor by applying power factor correction, basically a capacitor connected across an inductive load. This is normally done at the household level than for individual appliances such as refrigerators.
People normally don’t bother unless it has a real impact on the bottom line and they have big water pumps or large AC units operating regularly in big premises.
I mention it here because the two fridges I’m using for this solar project have very different power factors.
The new AEG fridge has a PF of 0.93 while the 10yearold Whirlpool has a PF of 0.60!
Basically it means the older refrigerator is drawing 35 to 40% more current than it should for its power rating and this extra current has to be supplied from solar.
So the first piece of advice when running refrigerators on solar panels is to use new units which are more likely to have a great power factor.
Older ones may have deteriorated to the point where they are pulling much more current than necessary.
How Many Amps Does A Refrigerator Use?
The electrical characteristics of the two refrigerators I compare are shown below:
Volts 
Run current 
Start Current 
Wattage (measured) W 
kWh (10 hours) 
Power Factor 

AEG 200 liter (63 liter freezer) 
230 
0.36 amps 
3.7 amps for 3 seconds 
78 watts 
0.38 kWh 
0.95 
Whirlpool 203 l. (60 l. freezer) 
230 
1.14 amps 
4.2 amps for 3 seconds 
134 watts 
0.52kWh 
0.60 
Measurement of the kWh over time is really the only way to determine how much solar energy is needed to run a fridge through the night, when the sun doesn’t shine at all, but there are other considerations.
How Many Amps Does A Refrigerator Compressor Draw?
From the table you can see that there are two measured currents – running and startup.
For most refrigerators, the compressor motor runs about 50% of the time, so the average current is much less than the fridge rated current. However, motors take several times more current when they start up.
It’s called ‘surge current‘. It doesn’t add to the overall energy usage in watthours, but it needs to be supplied when demanded.
If you had a solar panel supplying 5 amps maximum and a compressor motor started needing 6 amps, the motor would stall. It would try to start but would eventually blow a fuse or trip a circuitbreaker.
How to reduce fridge compressor startup current
One way of reducing appliance startup current is to connect a softstart device. This electronic circuit can be purchased on sites like Amazon and gradually increase the voltage to the motor so that it can’t pull the normal starting current.
It softstarts with a low current instead of hardstarting with maximum surge current.
The concept might be new to you because they are not used much in the home, simply because the loads are not normally big enough to warrant the cost.
A costanalysis is always necessary, but it’s hardly a consideration with normalsize appliances, only big water pumps and AC units.
How many batteries to run a refrigerator?
Even if a solar panel could handle the refrigerator running current, we have to consider the motor surge current. The best way to accommodate this is to add some electrical storage capacity to the supply i.e. a battery.
We can’t just use any old battery – it must be a deepcycle variety.
A normal leadacid auto car battery can deliver hundreds of amps to turn over a car engine, but they are very poor at delivering any substantial currents over a long period of time.
In fact, an auto battery shouldn’t be discharged more than 10 to 15%.
Even a deepcycle leadacid battery shouldn’t be discharged more than 50% before being recharged.
At this DOD (Depth of Discharge) a good quality battery will last around 500 charge/discharge cycles, so it pays to choose a battery that’s too big so that it discharges at a lesser rate, say 30 to 40%.
Best battery for 100w solar panel
The best option for any solar panel installation is the lithium iron phosphate type of battery (LiFeP04). They have a lot going for them, except for the price, which is higher than leadacid.
If on a budget, it’s very tempting to go for the cheaper option but over the longterm there’s really no contest – see the table below:
Lithium Phosphate 
Lead Acid 

Number charge/discharge cycles 
2000 (100% DOD) 
500 (50% DOD) 
Weight 
If LF = 10 kg … 
Leadacid = 30 Kg 
Deep cycle? 
Inherently deep cycle 
Special battery construction 
Cost comparision (lifetime) 
If LF battery = $100 … 
Lead acid battery = $233 
For most installations the LiFeP04 battery is a lifetime purchase, so when you consider the number of leadacid replacements needed every 5 or 6 years, LF is about 2.33 times cheaper.
They are also naturally deepcycle and can be charge/discharged 5000 times if DOD is kept at 80%.
In a large installation, weight is not really a consideration but for single appliance use, such as a refrigerator, such a battery may need to be mobile.
Lithium are about 3 times lighter, a fact I appreciate as I use solar on several projects where I need to lift batteries constantly (see my solar powered inflatable canoe project.)
For this test, I have two batteries available:
 Leadacid 12 volt 90Ah deepcycle ($250)
 Renogy LiFeP04 12 volt 50Ah ($449)
I’ll be hoping to use the lithium battery, but this will depend on the energy consumption tests. On the face of it the leadacid seems the better deal, simply because it has almost twice the amphour rating, but this is misleading.
First of all, deepcycle lead acid batteries are only discharged to about 50%, so that cuts the Ah capacity to 45Ah.
This figure will be reduced if the temperature goes up or if higher currents are drawn. It could easily happen that the effective capacity is only 35 to 40Ah.
LiFeP04 cells are inherently deepcycle – 80 to 90% DOD is common. The 50Ah model I purchased is rated at 90% depth of discharge with no damage, so they are very robust. Incidentally, the technology is very safe, unlike lithiumion.
On top of it all LiFeP04 batteries are a lifetime purchase if used at the 80% discharge rate.
Do I need an inverter for my battery and solar panel?
Unless you have a 12 volt DC refrigerator then you will definitely need to use an inverter. It’s the only way to convert the DC voltage from the solar panel/battery arrangement to the AC power that a household fridge needs.
What size inverter do I need to run a fridge?
According to my tests on two averagesized refrigerators, they run at about 150 watts with a startup surge of 5 times that, something like 750 watts peak power.
It’s always prudent to use power components that are a lot bigger than needed, just in case there are hidden peak currents or if you want to add to the circuits later.
I’m using a Modified Sine Wave 1000 watt solar inverter from a company called Solartronics, which is pretty cheap at under $100.
It isn’t the best quality but it will run a fridge (I hope)! The best inverters are pure sine wave.
They produce a really smooth sine wave that doesn’t interfere with any electronic frequencies like radios or hifi equipment.
Renogy Inverters
Old Refrigerator (Whirlpool) Energy Consumption/Battery Test:
Monitor the watthours consumed by a 200 liter with 63 liter integrated freezer Whirlpool Fridge/Freezer supplied by a 12 volt 90Ah deepcycle leadacid battery and an inverter.
For the first part of the test I wanted to see exactly how much energy the oldest refrigerator used over a 24 hour period. Not an estimate using averages and dutycycles, but actual readings.
The inverter was connected directly across the Varta leadacid battery and the fridge plugged in for 24 hours.
At this point there are no solar panels connected. I just wanted to see how long the fridge would last, or how much it used before I stopped the test. This way I will know exactly how much solar I need to keep the battery full.
The power factor is very poor at 0.56 lagging so I don’t have great hopes for this test. The wattmeter was already reading 1.42 kWh, so this needs to be deducted at the end of the test.
 After 7 hours running this fridge consumed 0.53 kWh of electricity which equates to 44Ah of 12 volt battery capacity.
At this point the compressor motor surge current of 3.8 amps was pulling the inverter voltage down to 226 volts (232 is normal in Europe) causing a lowvolt alarm to beep for several seconds.
The refrigerator carried on working but as the Ah used was almost 50% of the battery full capacity of 90Ah, it was a food time to stop the test.
Extrapolated over a 12 hour day with regular usage I would expect this appliance to use 75Ah, but I suspect the duty cycle is also very bad for this inefficient fridge.
The compressor cooling phase seemed very long compared to the new unit at around 16 to 20 minutes, which as probably due to a combination of inefficient/worn compressor parts, deteriorating thermostat and low power factor.
I gave up with this refrigerator as I considered it not worthwhile trying to run it with solar panels. On to the new AEG!
New Refrigerator (AEG) Energy Consumption/Battery Test:
Monitor the watthours consumed by a 200 liter with 6o liter integrated freezer AEG Energy Efficient Fridge/Freezer supplied by a 12V 100AH Smart Lithium Iron Phosphate Battery and a 1000 watt inverter.
This unit is the same physical size as the old Whirlpool, just newer and more efficient.
The running current is much lower while the compressor motor surge current is about 3.2 amps, so considerable higher as you would expect. Note that the power factor is over 0.9 – volts x amps is not far short of the measured watts at 83.6 VA.
The appliance consumed 0.19 kWh in 5 hours of normal use, which equates to 0.456 kWh over a normal 12 hour day, with and estimated duty cycle for compressor operation of 50%.
I think it’s safe to assume a much lower cycle through the night, so we can add 50% to complete the 24 hour period:
 Daytime energy consumed = 0.456 kWh (38 Ah)
 Night energy consumed = 0.456 x 0.5 = 0.228 kWh (19 Ah)
 Total energy over 24 hours = 0.684 kWh (57 Ah)
This appliance would run with the 90 Ah Varta leadacid battery used in the previous test, or with a 100 Ah LiFeP04 lithiumtype battery.
At a pinch it might be possible to use a 60 Ah lithium, but it could be a little tight, while 100% battery DOD would reduce the overall life.
The best option is the 100 Ah lithium, as the deepcycle leadacid would be discharged over 50%, which would substantially reduce its working life.
Whichever option we go for, 0.684 kWh (or 57 Ah) of battery capacity has to be supplied by solar panels.
How Much Power Does A 100 Watt Solar Panel Produce?
I almost exclusively use flexible solar panels. They’re powerful and very light – great for using on portable projects such as my inflatable kayak and DIY solar generator projects.
I’m going to use 2 flexible panels rated at 100 watts – see opposite.
The panels are built into aluminium frames to give them some rigidity, but even so they are 4 to 5 times lighter than a standard 100 watt solar panel.
I was a domestic solar panel installer in a previous life and for domestic installations I would invariably use rigid panels. For most DIY purposes, flexibles are perfect!
We need to find out how much energy one of these solar panels can produce in full sunlight, and also devise a means of measuring that power in watts.
Solar Panel Electrical Characteristics and Efficiency
Solar panels are sold and rated according to STC conditions (Standard Test Conditions) as follows:
 Irradiance (sun power) of 1000 watts per square meter
 Ambient temperature of 25 degrees C (77 F)
 Air mass coefficient of 1.5
Irradiance and ambient temperature are the two that impact solar panel performance in the practical world.
Firstly, your location at any particular time of day may or may not have an irradiance of 1000 watts per square meter.
That’s quite strong and equates to midday sun in most countries. Some countries may have far more than this while many will have less.
Secondly, for a fixed solar panel array the sun is not very strong in the mornings and afternoon. This is why solar installers multiply the solar panel rating by 4 hours of STC irradiance, as the average insolation (irradiance over time) for one day equals about 4 hours of midday sun power.
For these reasons, you’re never quite sure exactly what kind of power you’re getting out of a 100 watt solar panel. Add some cloudy weather and seasonal changes into the mix and it gets even more difficult.
How to determine solar panel wattage
This isn’t as easy as it sounds! A nominal 12 volt 100 watt solar panel has open circuit voltage Voc of about 22 volts and short circuit current Isc of 5.92 amps – it’s OK to short the leads together to test this current, by the way.
When you short out a solar panel, the voltage drops to zero. It just so happens that the maximum power a 100 panel produces occurs when the current is 5.41 (about 95% of Isc) and 18.5 volts:
 18.5 x 5.41 = 100.085 watts
Make sense? But is this what you get? Let’s find out.
Solar panel testing equipment
You can easily test solar panel Voc and Isc with a multimeter (see below), but this basically just tells us that the panel is functional and it can deliver current.
It doesn’t really tell us how much power it can generate. We need another method to measure how many watts this panel is capable of.
A better arrangement would be to connect the 100 watt solar panel through an MPPT solar charge controller to a partlydischarge battery that also feeds a load.
Renogy Solar Charge Controllers
A solar panel delivers its maximum power for any irradiance value at the Maximum Power Point. For my 100 watt panel this occurs when 5.41 amps flows at 18.5 volt.
We need a solar charge regulator to govern solar panel output and we have a choice of 2 types:
 PWM (Pulse Width Modulation)
 MPPT (Maximum Power Point Tracking)
PWM vs MPPT Test
PWM pulls the solar panel output voltage down to whatever value is required by the battery, while MPPT regulates the voltage so that maximum wattage is always exploited – MPPT is the one we want.
Testing the 100 watt solar panel output power
The battery will be partly discharged but I need to make sure it draws maximum current from the panel, so I connected a DC load in the form of a cupsize water heater.
This is rated at 140 watts so pulls about 11 amps from the battery which is partly replenished by the MPPT controller.
The DC wattmeter only measures the power coming from the solar panel, and also records watthours (Wh) and amphours (Ah).
100 Watt Solar Panel Instantaneous Power Measurement Results
The tests were carried out in March in the South of Portugal, so plenty of sunshine but not yet strong as it’s barely Spring.
 the 100 watt panel output 1.9 amps and power of 25 watts
Two things are immediately apparent:
 Instantaneous measurements tell us very little, due to weather changes, clouds, time of day,etc
 Seasonal changes in power output are going to be huge, making it difficult to size the solar power needed across all seasons.
We need to know the watts required over time and how it fluctuates throughout the year – it’s no good if 2 x 100 watt panels can power the refrigerator in the summer but not in the offseasons, for example.
Use the calculator below to find out how much average energy a 100 watt solar panel will generate in your location. Use the link in the calculator to find the ‘peak sun hours’ for you area.
This is a historical estimation of the number of hours per day during which the irradiance is equal ton 1000 watts/m2 i.e. the STC standard irradiance which equates to maximum panel power output.
In my location, peak sun hours are today (March 4) 4.8 kWh/m2/day so the panel is going to deliver about 384 watthours a day, which is considerably less than the 684 watts previously estimated as the fridge’s average energy consumption for 24 hours.
Can a 100 watt solar panel run a refrigerator?
So now we know that a 100 watt solar panel is just not going to be enough, based on average daily production over the year.
I’m going to add another panel so we have 200 watts available connected in series to the MPPT controller.
Inputting 200 watts into the calculator we see that now the output is 768 watthours per day, so more than enough to run the fridge.
But what about seasonal variations? We need to know how two panels will perform during the Winter.
Using another site for monthly insolation I laid out the peak sun hours across the year – see below:
Month – Location: Algarve, Portugal 
Peak Sun Hours (kWh/m2/day) 
January 
2.99 
February 
3.28 
March 
4.80 
April 
5.79 
May 
5.93 
June 
6.69 
July 
6.94 
August 
6.74 
September 
5.44 
October 
4.14 
November 
2.65 
December 
2.35 
I can now see that using the peak sun hours value for August in the calculator 200 watts of solar will generate over 1110 watts – more than enough! Now let’s check out the worst month, December.
Inputting December’s peak sun hours value of 2.35 kWh/m2/day we get 376 watthours, which is not enough to run the fridge. 300 watts of solar is still not enough at 564 watthours output per day.
We need to add yet another panel to bring the total to 400 watts to be sure of running the refrigerator throughout the year. This means our system would generate a huge 222o watthours per day in July!
How Many Solar Panels To Run A Refrigerator?
In conclusion, we could say that even in my very sunny location in July a 100 watt solar panel is not going to run a refrigerator. Two solar panels rated at 100 watts will run it for most of the summer.
Most locations will need four 100 watt solar panels to operate the average fridge during winter. The exact number and ratings can be extrapolated using the calculator and solar irradiance sites in this post.
How To Increase Solar Panel Output
There’s a lot of research going on at the moment to try and improve the output from standard solar panels.
It’s become a kind of Holy Grail for the DIY fraternity but it’s met with limited success. It remains for the individual to work out if it’s costeffective for him/her.
How to increase solar panel efficiency using solar panel mirror concentrator
In a previous post I setup a 100 watt solar panel with a mirror and measured an increase in output watts at about 30%, which was obviously a great result.
At that time I didn’t measure the temperature rise over time, which could have reduced the output as output falls by 0.5 watts per degree rise in C.
Nevertheless, it’s wellworth a try for the determined offgridder, if the cost is not too great.
Improve panel output with an automatic solar tracker
A more proven method of increasing solar panel output throughout the day is to automatically track the sun’s movement.
Single axis tracking tracks the sun’s horizontal travel, which is the greatest movement and can give up to 40% greater output watts.
Double axis track horizontal and vertical movements and can give up 65% more power. It has to be said that these commercial systems are expensive and are often not costeffective for small systems.
That said, DIY tracking systems are easily constructed and can make a difference.
If solar cells keep reducing in price at the same rate as the past 10 years, domestic installers should think carefully before paying for a tracking system – it’s often cheaper to simply add a couple more solar panels.
(More details of PWM vs MPPT comparison)
Related Questions:
Can a 300 watt solar panel run a refrigerator?
In most locations 300 watts of solar power would not be quite enough to run a fridge all year round. It would be adequate in summer, but wouldn’t produce enough energy for the Winter months. (see post)
What size charge controller for 400w solar panel?
A typical 400 watt solar panel outputs about 8 amps at 6o volts. An MPPT solar charge controller may be used to charge batteries from 12 volts to 48 volts and will generate extra current as the voltage is reduced.
Discover your solar saving potential
This means that an MPPT controller could supply 32 amps of charging current and so a 40 amp controller is recommended.
Can you run air conditioner off solar panels?
Running an AC unit ‘offgrid’ with solar panels is not a costeffective proposition. As much as 3kW of solar panels is need for average solar powered air conditioner.
It is viable if wholehouse gridtie solar is installed. In this way the grid can supplement the power needs and the solar power can run the rest of the appliances during the months when AC is not needed.
Resources
https://www.nrel.gov/docs/fy01osti/26468.pdf
EREC Brief_ Energy Use of Some Typical Home Appliances