How To Power Through A Heatwave – In Comfort – On 100% Solar Energy

solar power and heatwaves

Is it possible to keep a family home comfortable in an extreme heatwave using only solar energy? Yes.

Last week I received a phone call from an ABC journalist. Unfortunately he didn’t want to invite me on Gruen to talk about the joys of solar advertising, or book Ronald & Tonto23 to perform on Mad As Hell. Rather, he’d been told that solar panels lose efficiency when it gets hot, and was keen to learn more. Specifically, he wondered how much solar panel efficiency would be lost if the temperature hit 45º during the heatwave – as forecast for the following day.

I spent the next 20 minutes going through the 3 basics of solar power and temperature that, in my experience, cause confusion to non solar-nerds:

  • How solar panels love light, but dislike heat.
  • The difference between solar thermal and solar PV panels.
  • Why the temperature used to determine the efficiency loss in a solar panel is the panel temperature, not the ambient air temperature.

After the call I worried I may have inadvertently given the impression solar panels cannot be relied upon to power your home (or help power the nation) through extreme heatwaves. I really don’t want the anti-renewables brigade to think they’ve discovered solar power’s kryptonite: extreme heat.  Hence this blog post showing real data from two rooftop solar systems in Adelaide that reliably soldiered on through the oppressive heat.

The next day, as forecast, we experienced the extreme heatwave in my hometown of Adelaide, with the top temperature of 46.6ºC breaking all records. I have a relatively efficient solar and battery powered home. I also have monitoring on a less thermally-efficient, solar powered (no battery) home down the road, so I thought I’d share how both houses performed on the hottest day Adelaide has ever seen.

And it really was pretty fricking hot. I rode to the shops (to get a red onion) at 3pm and it was like pedaling through a wall of hairdryers. I didn’t see any other bicyclists (or pedestrians) on my 400m round trip.

For the record, these were the half-hourly temperatures in my part of Adelaide on the day in question (24 Jan 2019):

House #1 – My home: 6kW of solar, 13.5kWh of battery, split aircon, good thermal envelope.

Q. How much energy did I need to run the house and keep everyone cool?

A. 39.5kWh

This is the 5 minute power consumption of my house on that day. The total energy used1 is a whopping 39.5kWh.

My home’s typical daily consumption is around 15kWh. That powers a family of 5 and a small business. So you can see that keeping cool on very hot days takes an exceptional amount of energy. Even in a house with a good thermal envelope and a modern, efficient air conditioner.

Q. How much solar energy did I generate?

A. 34.8kWh

A typical solar panel loses 0.5% of its efficiency for every degree over 25ºC that the panel’s temperature is at. In 45ºC I’d expect the panel temperature to be close to 70ºC. So I’ll be losing about (70-25) x 0.5  = 22.5% of my power from the crazy-high temperature.

I have a 6kW system so I’d expect to lose about 1.3kW of power due to the heat and perhaps another 5% (0.3kW) due to other system losses.  i.e I’d expect the peak power of my 6kW system to be about 4.4kW on a 45º day.

Here’s the solar electricity generation laid over the consumption:

Two things to note:

  1. My 6 kW system peaked at 4.4 kW as predicted.
  2. Even if I didn’t have a battery, the solar alone covers most of my consumption over 24 hours. Read on to discover how much.

Q. How much of my solar electricity was self-consumed (ignoring the battery)

A. 22kWh

The red area here shows the directly self consumed solar energy:

So from 39.5kWh consumed, 22 was supplied directly from solar electricity.

Q. How much money did my solar save me over 24 hours?

A. $10.92

The red, self-consumed solar electricity saved me (22 x $0.38) = $8.36 because I didn’t have to import that energy from the grid.

If I didn’t have a battery, the remaining 17.5kWh would be simply topped up from the grid.

The cost of that imported electricity would have been $6.65. But that would be reduced by the $2.56 earned from exporting the excess solar electricity at 20c per kWh.

So, with solar power only (no battery) my energy bill for the 24 hours would have been $4.09, compared to $15.01 if I had no solar panels.

My solar alone saved me $10.92 over the 24 hour period.

But, I have a Powerwall 2, 13.5kWh battery on this house.

Q. How much of the remaining consumption was supplied by the battery?

A. All of it.

The battery was happily discharging before sunrise and started to recharge from the excess solar electricity at about 8:30am. The battery didn’t run out until 1am the following morning.

The battery allowed me to self-consume 17.5kWh of solar energy instead of exporting it to the grid. The net savings per kWh discharged from the battery are calculated as the cost of grid electricity minus my feed in tariff ($0.38 – $0.20) = $0.18 per kWh.

The battery saved me (22kWh x $0.18) = $3.96

Compare that to the solar-only savings of $10.92. And consider that a really good 6kW solar system costs about $7,000 installed whereas a Powerwall 2 costs $15,300 fully installed.

Some things to note:

This was an exceptional day. The battery was used at 130% capacity. Normally I use about 5kWh (36% of battery capacity) giving me typical daily savings of $0.90 from the Powerwall.

I didn’t use any grid energy that day. That’s pretty cool. I didn’t place any load on a very stressed grid2.

Here’s the full energy profile for the 24 hours courtesy of Solar Analytics:

power profile

The negative blue area is the battery discharging to offset consumption.

House #2: 6.6KW solar (east/west split), ducted aircon, poor thermal envelope.

The guys in this house (AirBnB guests) used a phenomenal 70.5 kWh of energy, of which 65.3 kWh was the air conditioner. The air conditioner is about 13 years old and ducted.

The house generated 35.9 kWh of solar electricity with its 6.6kW of Winaico solar panels – half east and half west.

The residents self consumed 30kWh of solar energy – saving $11.40 and reducing the electricity bill for the day from $26.80 to $15.40.

Adding a 13.5 kWh battery would have saved between $1.06 and $3.33 depending on how much charge it had at the start of the 24h period (0%-100%) and assuming only solar electricity was used to recharge the battery3.

Can a modest solar system plus battery power you through a heatwave?

If you have a home with good thermal performance, yes. You can get through a heatwave without spending a penny on grid electricity4. The battery is a luxury though – as the vast majority of the savings will be from the solar panels.

If you live in a typical Australian home with ducted air conditioning and a poor thermal envelope, you are going to need a huge battery and big solar panel system to be 100% solar powered in a heatwave.  House #2 above would need 3 Powerwalls and a 15kW solar system. That’s at least $50,000 of gear. If you want to reduce such a home’s impact on the grid on extreme days, then the most cost effective way to do that is to improve the thermal efficiency of your home (gaps, glazing, insulation) and get the most efficient air conditioner you can. It’s not as sexy as a battery, but it will make your home more comfortable all year round and save more money.

If you combine that with a reasonably sized solar power system, your home can be transformed from being part of the problem to becoming part of the solution when these increasingly common heatwaves hit. And I think that’s a noble endeavour.


  1. mathematically, the area under the line
  2. but did use some from 1am to just after sunrise the following day
  3. Because I’m usually accused of being mean to batteries, I’m very generously assuming the battery could start off as fully charged at midnight.
  4. Although I consumed about 5kWh from the grid the following morning as the battery emptied by 1am, the cost of those kWhs was more than offset by all the energy I exported the following day.
About Finn Peacock

I'm a Chartered Electrical Engineer, Solar and Energy Efficiency nut, dad, and founder of My last "real job" was working for the CSIRO in their renewable energy division.


  1. This presumes the heat is a dry heat and not a humid cloudy day. We can have 37 degree days and be making virtually no PV, I hate those days

  2. Hi Finn,
    question on the PowerWall2.
    If it cost in the order of $15,000 to install, and the ‘expected lifetime’ is around 10 years. doesn’t that mean you need to recover at least $4 a(and some change) to recoup the costs?
    As best I can see in your “worst case / best case” scenario above, you nearly “broke even” for that single day.
    I have a 4kW system, and most days I export some excess back to the grid (NSW – approx $70 a quarter) but I am still having trouble justifying a battery system. I would _love_ to be “blackout proof” (well, “blackout resistant”) but the economics at this stage are still not quite there.
    love the updates!

    • At $15,000 it is almost impossible to make the Powerwall pay for itself in 10 years. You’d have to cycle it multiple times per day and be on a Time-Of-Use tariff with very high peak charges.

    • Yes but your assuming the PW2 will just suddenlecstop working after 10 years. While that is all the warranty maintains you could potentially double that lifespan. Finn – have you got figures on degradation of your PW2 yet?

      • Real data would be good to get idea of PW2 life estimates. But in truth, given that it is only 2 years old, there is probably not much useful data available that will give anyone a real idea of expected lifetimes (unless the data is very bad while I am not aware of yet), Typically lithium ion batteries tend to maintain pretty good capacities (even sometimes improving after break in period), usually with a quite slow decline, until they fall off a cliff. 70-80% is not uncommon for the rough timing of when they quite commonly fall off the cliff (different batteries even within the same chemistry class will have different characteristics). Tesla would be very aware of this when they choose their warranty. Batteries with no degradation is the holly grail may people have been working on for over 100 years and the improvements have been slow.

        The best scientific “Data” I have seen on lithium ion home storage battery performance is here :-

        The process is far from perfect, but it is at least analytical. The challenge is they have not been able to control the PW2 well enough to cycle it. But the concerning thing for me is that the PW1 is not performing well at all. Unfortunately as best I can understand it is using the exact same batteries as the PW2, and as far as I can tell even Tesla are not claiming changes in the PW2 should lead to longer battery life or suggested it might be so with an improved warranty. So I think it would be reasonable to expect similar degradation. From the numbers there, it looks like it might struggle to get 50% of the way through the warranty. In its defense, for technical reasons, they are having to cycle it at a faster “C” rate than the other batteries in the test. C/2 compared to C/3 for the other batteries. But it should also be noted this is a rate that is supported in real world deployments so it should be able to handle this. This is likely to be having a negative impact on performance when compared to other batteries (damage due to higher discharge rate, and also potentially heat which can be problematic). But to me the negative impact looks out of proportion to the higher charge/discharge rate, and given it is not beyond it specification is concerning to say the least.

        But based on all the Data I have seen, and based on the type of chemistry they are using, and all we know about that type of chemistry, I think you would be very optimistic to assume you are likely to get 20 years out of the battery. I personally suspect you should not count on much more life outside of warranty. When calculating economics for my own purposes I generally only assume warranty, and then do what I think it more likely to be a realistic best case scenario that is 50% more than the warranty. But at this stage we are all only guessing. I suspect your best chance of getting 20 years if they choose to replace a close to end of warranty battery with a brand new 1. It might be Tesla know this, but are happy to create confidence on the product with a decent warranty, and sell more and wear the warranty exposure and get the benefit of collecting the data so they can develop and improve the special sauce to one day hopefully develop a battery that will be good well beyond 10 years.

        Now don’t be distracted with people pointing at Tesla cars and looking at battery degradation there and all sort of optimistic bumpkin. It is such a different use case, as I suspect to be irrelevant. eg car batteries are not generally cycled to 90% DoD every day, and we know heavy cycling of lithium ion batteries is hard on them. Finally while you have all sorts of experts of forums etc suggesting all sorts of things, as near as I can tell neither Tesla or people like Jeff Dahn (NMC battery pioneer / academic who has got funding from Tesla for some of his work) are suggesting this chemistry should be good for anything significantly longer than the warranty. Though they are all working on it, and in time home to be able to address the reasons for battery degradation to be able to increase their life.

        • Hey Matthew,

          Great insights. Time will tell. Interesting that Tesla provide no easy way of measuring the remaining capacity of the battery.


          • I don’t yet have a PW2 (I plan to get one for the blackout protection benefits rather than the economics that don’t provide anything near a sensible payback on economics alone for me…I have also been waiting in vain for better 3 phase support, as I want to keep solar up on more than 1 phase, and support for generator which was promised, but near as I can tell has completely disappeared from the radar in any sensible and useful way). So can’t comment with any certainty on the ability to measure remaining capacity.

            However, I have seen people get figures showing the current capacity of the PW2 on a whirlpool forum, so pretty sure it is possible (or at least was possible). It might have been via the web interface or API, but I did not get the impression it was hard to do. From memory there were a few people who compared capacity when they brought it, to capacity now. And the numbers looked pretty good, with no alarming decline in performance. However at the time most people were less than a year old, so would not expect much or any decline anyway. So probably too early to make any conclusions.

            It is certainly be interesting to see how it unfolds.

          • Ronald Brakels says:

            Hi Matthew, Ronald here.

            If you are just looking at a Powerwall 2 for blackout production I suggest looking at a generator instead as, unless you have particular reasons for preferring a battery system, it is likely to be much more cost effective.

          • Hi Ronald, I already have a generator, but I have lots of reasons I want to add a battery :-
            1. reality is they are noisy and smelly, a problem which a battery solves.
            2. I also don’t really have any suitable space I want to permanently set it up.
            3. because I run IT business and servers, I also need UPS functionality. Sure, I can spend money on UPS with big enough batteries to keep me running till I can manually start the battery, but a PW2 fills some of these requirements.
            4. We have a LOT of blackouts here. Averaging well more than a dozen blackouts a year, and some of these last many days.

            One of the reasons I don’t have PW2 yet, is in June 2017 Tesla sales person said there would be a gateway that supported backup and batteries on multiple phases, and generator input. Happy days, and based on this my deposit went down. However, unfortunately neither of those things have materialised. If I had that, I would have 2 x PW2, on 2 phases, and with that I would have plenty of backup power from the battery, be able to keep 2 phases of solar running and charging the battery which will give me long blackout protection, and when I notice battery is running down in a blackout (maybe because of high power demand or low solar output), I will be able to start the generator and run it for a few hours to recharge the batteries to carry me through to the next day. All of this with avoiding having to run a generator 24×7 which as anyone who has actually lived of generator power, listening to the hum of the generator 24×7 can get tiring. That was the dream. But until Tesla (or someone else at a reasonable price) can deliver that…it is still just a dream.

  3. Richard Maddever says:

    You don’t mention your air conditioner settings. Appreciable difference in energy probably needed between 20 deg and 25 deg a/c setting.
    Some hot days can be quite cloudy, as we had quite recently, and solar panel output was appreciably affected.

  4. At my place the panels face NE,like the house,so when the sun is blasting in from the west the solar power has gone,and I am wondering whether anyone is doing special panels that could face west. I am thinking of a system where you crank them around to follow the sun. I have seen big installations, but perhaps 2_3 panels might supply the aircon.

    • Des Scahill says:

      Hi Robyn,

      You haven’t given any details but I live in QLD so that might be some help. My house, which is a lowset, and has a colourbond gabled roof, faces almost directly East. That wasn’t really a problem for me, because the NE side still provided ample space. and there was ample gable surfaces facing in all directions which provided some flexibility in locating panels.

      The near-optimum split in my case was to put 55% of the panels facing North, and 45% facing West. That had the overall effect of ‘spreading’ the solar power generation over a longer period, and, although the TOTAL generation actually drops, that ‘loss’ is more than compensated for by the extension of time. ie – a lower overall production on any one day gets offset by the fact that I get a longer period of time for useful self-consumption. So, if for example in summer we start cooking tea at say 3:30 PM, there’s a very good chance that most days there will be enough generation to completely cover the draw-down by a 2600 watt hot plate.

      So It might be worth your while to simply look at moving roughly half your existing panels so they face West as one alternative to consider. There’s a few variables involved, so do your sums very carefully, and compare that to changing your existing consumption time patterns

      From an installers point of view, he’s effectively being asked to ‘de-install’ X number of panels, and ‘re-install’ those same panels elsewhere, along with some re-wiring to because of increased connection lengths. As well, might need to get ‘approval’ once completed. Good idea to talk to your electricity retailer about what you propose before actually doing too much.

    • Erik Christiansen says:

      Very good info here. I’d planned to put a 7 kW array on my 30 degree directly north facing roof (not quite up yet), but to lose power to the aircon in mid-afternoon (I’m off-grid) would be ghastly on a 45 degree day. It seems it’d be much better to put half of the panels on the west facing roof of the old house, to power that aircon through the hot afternoon. Then I can hammer the poor battery after sunset, if necessary. (Only planning on 10.4 kWh, and running the generator flat out to power aircon would have to be with biofuel, or I’d be part of the global aircon death spiral.)

      A DC coupled system would need a very big battery bank to allow the MPPTs to feed a lot of power to the battery inverter, I figure. Going to an AC coupled system, with PV inverter and battery inverter/charger, makes it much more feasible to directly utilise high levels of array power without exceeding the charge rate limits on a modest battery bank. Would need a PV inverter with two MPPTs, one for each array.

  5. The numbers look good if your into these things. Ill keep on pestering my wife with the daily calculations for a bit longer till she throws something at me

  6. Maria Gardiner says:


    I too hail from hot, lovely Adelaide.

    I found I was able to export 11.5kw that day and used nothing from the grid. I have one 5kwh system and one 6kwh system and the 13.5kwh Tesla Powerwall 2 (so I should be able to manage without drawing from the grid!). However, when I moved into my big house a few years ago i was stupidly talked into installing a big ducted reverse cycle system (worse decision ever – go with modern splits instead) and it runs at about 4-5kwh per hour – it rarely drops below that and never did on that hot day. So the only way I kept the balance was by watching my monitoring system. I also wanted to keep around 40% in my battery in case there was a blackout – I was keen to justify paying the extra $800 for backup capability after the big blackout scandal!

    So all of that is by way of saying I was still able to come out ahead despite having a bad air con system and thermally inefficient house, but it took a bit of monitoring and what some might describe as a sh*t tonne of solar. But it was good fun watching it all unfold on the monitoring system!

    My daughter runs exactly Finn’s set up (encouraged by good ol’ Mum) and she came out with drawing around 20kw from the grid after running 3 air cons for pretty much the 24 hour period – so more like an airbnb user! But still a lot better than if she did not have solar. She self consumed every bit of it and only just filled her battery (I might have made her turn off her air con for a bit).

    Boy I love solar (and only have batteries because I managed to happen by on the one occasion AGL have ever been generous in their entire lives!)

  7. Patrick Comerford says:

    There is also one other important consideration in your homes solar and battery performance. If you had been one of those unfortunate grid customers who had their power shut off or lost during that period which I believe was several tens of thousands, then you would have been able to continue supplying power to your house ac and other essential consumers without even noticing it had happened until you received your phone notification from your Tesla Powerwall.

  8. Good analysis, thanks.
    Just wondering, with your Powerwall 2, if you have a blackout, do your panels keep charging your battery and powering your house ?
    I know you can use battery power in a blackout.

  9. Terry O'Leary says:

    Hello Finn

    Nice analysis of solar power for Air conditioning.

    We are currently in the process of replacing our 10 year 2kW solar panels on our house (and SMA inverter) with a new Fronius inverter and 6.25kW of Winiaico panels.

    So unfortunately we were down to our 18 month old 5kW of Winiaico panels (with Fronius inverter) and our Selectronics SP Pro inverter with 10 kWh of BYD batteries fitted in & on our garage. The hot weather hit while we are minus our old 2kW set up, but not yet had got our new 6.25kW.

    We have been really pleased the way our garage mounted 5kW system powered our house and split system air conditioner during the heat wave. We have an energy efficient house in NSW with 2 people at 800m, 20 km east of Canberra. We have an open plan kitchen dining area and directly living room & main bedroom. We had set our efficient Daikin 8kW split system air conditioner at 24C which keeps those 4 rooms at 25C while using only about 1kW-1.2kW when outside is in mid to high 30’s.

    During the late afternoon earlier in the week at around 4pm with a thunderstorm approaching when we checked our solar system data. We noted we were at a comfortable 25C with 34C outside and we had exported 12kWh to the Grid in that day. Our Battery was fully charged (100%) and so ready for any power disruption from the storm. All this was from only 5kW of solar panels.

    It will even be better when we get our additional 6.2kW of panels. All we will need then is a reasonably priced EV to charge from our excess solar production.
    Solar powered home with air conditioning works well.

  10. Robert Leahy says:

    Excellent analysis Finn. What sort of data collection and gear do you use?

  11. I was thinking about this the other day.
    There are many of us frugal users who normally export a lot to the grid. My ratio is about (2-15) 2kWh import to 15kWh export this time of year.

    Leaving the storage and heat losses aside, with temperatures this high even us frugal users are reaching for the AC remote. My ratio during the two January heatwaves was around 1-1. Our fans got a lot of use too.

    So, the grid is missing out on a lot of solar exports.
    Solution. More solar, preferably some Western facing and some misters on RCACs would help.

  12. The answer to that is “it depends”. Some of the different answers :-
    1. If you only have single phase solar, and the battery is installed on the same phase as the solar, and the battery was installed to support it with the backup gateway, then the answer is yes.

    3 phase homes answers get a LOT more complicated :-
    1. If you have 3 phase solar inverter, then the answer is NO, and probably always will be.
    2. If you have an old micro system (ie Enphase) on 3 phases, and you have battery setup to support it, then the answer is probably yes.
    3. If you have a new micro system (post about Oct 2017) (eg Enphase), then rules in AS4777.2015 probably require that they shutdown in a blackout. You might be able to get your installer to give you a profile that enables it to work, but this will probably not happen by default, and they might not allow you to do it.
    4. In theory the AS4777.2015 rules that cause an issue with support of solar in a blackout that affect new micro systems, technically also apply to single phase inverters connected to homes with 3 phase power. But my guess is that in most cases people with single phase solar, and setup to support this in a blackout should be ok.

    Sorry the answer is not a simple one. But until the AS4777 regulation catch up with requirement to keep solar running in a blackout, this will always be a murky area.

  13. phil colbourn says:

    can you tell us how how big is your air-conditioned area? eer of your a/c? and perhaps perimeter of air conditioned area?

    • Hello Phil

      Not sure who you are asking question of, but can tell you about our set up.
      Our house has an insulated waffel concrete slab floor, is orientated E-W, no windows on the west side, veranda on south, pergola with some shading to north. It came second in SE NSW HIA energy efficency awards in 2000 when it was built.
      The ceiling is insulated with R4 batts and sarking, internal walls are all insulated~R1.5, South facing galss double glazed and all window and doors have heat reflective roller blinds. The airconditioned area is 8m x 16m and has rooms we do not air condition to the East and West. The area can be air conditioned by about a 5kW split system, but we went with an 8kW system, which will be not working as hard. We placed the condensor at the Eastern end of the house in the shade.
      We also thermally manage our house and at night open the sliding galss doors to cool down the thermal mass of our house, and close up as day warms up and do not turn air conditioning on till around lunch time or if it gets to 25C inside our house.
      The EER of our unit is about 3.5 and as the unit is not working flat out and the condensor is well positioned its efficiency should be better. I think COP/ EER is specified at maxium output (not sure). However, we try to make it easy for our air conditioning to do its job. It heats well in winter, but mostly we use our wood fire.
      We have tried to make our home very comfortable, sustainable as possible and energy efficent. So far it is working very well. We use only rainwater. We are still connected to the grid as we can export our excess solar, we use a lot of our own wood for heating and prodcue more electricity than we consume. It is very possible to live very comfortably with modern technology like air conditioning and yet be very sustainable.

  14. Down south (near Hindmarsh island) the temperature was also unbearably hot. I’m connected to a poor transformer feed and the voltage will often be around 250 volts or higher and rarely goes under 240 volts so I suffer from Inverter ramping issues all the time.

    I have a 6.48kW array (East/West) and normally on hot days means lots of electricity is being consumed in the area which brings the grid voltage down below the magical 250 volt mark where ramping starts.

    On the 28th January though the voltage was at one point way down to around 230 volts which is rare for this area and it never reached 250 volts so I was expecting to do well with my generation but when observing my graph and expecting to see a 5kw peak graph limit it was instead showing between 4 and 4.5 kW for the whole day and with no cloud about.

    I knew solar panels do not like the heat but was surprised to see such a large reduction in power on that day. I have a Sunnyboy Inverter (Chinese built) which performs well normally and is not located in direct sunlight but it was extremely hot to touch so whether the Inverter was reducing it’s output level to keep it’s internal temperature within limits I don’t know.

    I only managed 38.16 kW for the day but still managed to pass my expected daily average for January.

  15. Daniel Huppert says:


    I think you might have calculated the additional saving wrong mate.

    “Adding a 13.5 kWh battery would have saved only another $3.33, assuming they only used solar electricity to recharge the battery”…

    I am assuming you are working on a rate of about $0.246/kWh to calculate this figure, which is about the standard rate. What I think you have failed to factor in is the fact that if 13.5kWh was saved in the battery for future use (in reality, you’d have to feed closer to 15kWh into the battery to return 13.5kWh later on), that 15kWh would not have been fed into the grid, and thus would have missed out on at least $1.65 worth of feed in credit (assuming about 11c feed in rate).

    Therefore, having a Tesla battery would have only saved an additional $1.68, not the $3.33 you claim, as additional savings above and beyond what was safe via self consumption and offset by feed in.

    The only way you could have possibly saved $3.33, would be to assume you were paying a $0.368/kWh peak rate (being the effective rate after all discounts applied, and assuming an 11c feed in tariff, and assuming 15kWh was fed into the battery to discharge 13.5kWh later), and if that’s the case, whoever is paying so high a peak rate is being ripped off by the energy retailers.

    I have seen that many people in solar and battery sales always tend to over inflate the savings possibly from batteries, because they neglect to take into account that the savings will always only need the difference between what you would have got from feeding that power into the grid, vs what you save by storing it and using it later on.

    In most cases, the difference amounts to about 10c/kWh when all is factored in….

    • Hi Daniel,

      Did you read the rest of the post or only that paragraph?

      I really don’t think you can accuse me of over inflating the payback of batteries.

      This blog was the first place to explain how you must subtract feed in tariffs from consumption tariffs to calculate payback.

      It was the first to expose blended payback.

      It was the first to point out the negative CO2 outcomes of home batteries.

      I get so much stick from the establishment solar industry for simply calculating battery payback correctly you would not believe.

      In the post I actually spell out the calculations of the battery payback on my home as:

      “The net savings per kWh discharged from the battery are calculated as the cost of grid electricity minus my feed in tariff ($0.38 – $0.20) = $0.18 per kWh.”

      I point out that my typical powerwall savings are 90c per day.

      For House #2 I am looking at a 24hr period. If the battery is fully charged at midnight, then recharges with spare solar it will discharge 18.5kWh. 18.5 x (0.38 – 0.020) x 18.5 = $3.33.

      I’m trying to be really generous with my assumptions to show that even if the battery was fully charged at midnight the savings are still small compared to the solar. I’m being generous because 99% of the time I’m told that I’m ‘anti-battery’ and should assume > 1 cycle per day.

      (But point taken – I’ve edited the post say that I’m very generously assuming a fully charged battery at midnight)

      P.S. I pay 0.38 per kWh consumed after all discounts and get a 0.20c FiT. Retailer plan was chosen to optimise final bills using this tool:

      Both houses get negative bills. I’m hardy being ripped off by the retailers.

      • Daniel Huppert says:

        Hi Finn,

        My bad, wrote that at 1am in the morning, for some reason I did miss that paragraph. Also, I forget you are from South Australia, goodness the price of power out there is bloody high, I feel for those, especially pensioners, who don’t have solar power…

        Only thing to factor in to be completely accurate is to add in the charge/ discharge cycle inefficiencies. I have seen figures which show as high as a 20% loss when comparing kWh in vs kWh out of a battery… This figure also encompasses the built in battery management system constantly drawing power as well.

    • Maria Gardiner says:

      I live in SA (where Finn lives). We pay 41c KWh (including gst) for our power.

    • Welcome to Adelaide where 36c is a normal price and not just peak, that’s 24h a day. I do get 20c feed in tariff to the grid and if I didn’t care about that then I could get slightly cheaper, probably around 33c, but overall these rates are normal.

      But even so, $3.33 does seem like too much of a saving for the second house. There was only 5.9kwh to recharge the battery during the day and based on the overnight usage I wouldn’t have assumed that there would be anything left in the battery after midnight. So 5.9kwh to recharge the battery forgoes $1.18 of feed in at 20c and let’s you get about 5.3kwh out of the battery which saves $1.90 at 36c for a net saving of 72c.

      • Agreed – I’m overreacting to the critics by assuming 100% charge at midnight – when it is more likely close to empty. I’m just trying to show that even on a perfect day for battery capacity utilisation the savings are dwarfed by those from solar. I’ve edited the post to give a range of battery savings based on the charge of the battery at midnight.

  16. geoffrey redmond says:

    id rather have a generator out the back. very economical and starts every time whether the sun is shining or not. and by the way where do you put the batteries for your solar systems , inder the bed.

  17. WOW! … I just sat in a bathtub full of cold water ~ next to the esky ~ and watched the racing on TV., and let the panels look after themselves!

  18. Dennis Murphy says:

    Finn, I have so enjoyed reading this analysis. And all those of other users with analytical aptitude ..Thanks
    I’m going to expand my small system, forget batteries, and extend my Reverse Cycle AC using a 2nd split system unit. and Suffer the kitschness.

  19. Gregory John Olsen says:

    Thanx for your great treatise, Finn. You provide great evidence for the fact that, currently, it’s uneconomical to install a battery, Maxing out on array size, a minimum of 6.6kW of panels with a 5kW inverter, would do the trick in most cases. 🙂

    BTW, I live on the NSW Central Coast and we have two PV systems totalling 7kW of panels and 6.5kW inverters. Last month we imported 2.75kW/d and ended up with a $143 credit from AGL Solar Savers. No battery whatsoever. 🙂

  20. Hi Finn,
    Thanks for all the details, I was interested to see how we did on our really hot days. We had one air con running all day, turned on the other in the afternoon for a few hours, and I didn’t charge the electric car until the cooler days, and we definitely at least broke even.
    However, one thing I noticed is that the air cons were really struggling, never cycling on and off, just pulling 3kW constantly. I’ve had one sparky (the guy who installed the solar) tell me they were good units ten years ago, but they’re pretty inefficient now.
    Any truth in that? Should I consider replacing perfectly functional air con units just to decrease my reliance on filthy filthy coal?

    • Ronald Brakels says:

      Hi Mark, Ronald here.

      Air conditioners will decline a little in efficiency as they age, but what your installer was probably mostly referring to was new units can be perhaps 25% more efficient than 10 year old ones. If you know the energy star rating of your air conditioners or can find it online, you can compare it to new units to see how much energy they can save you. If you average 3 kilowatt-hours of air conditioning electricity consumption per day and replace them with units that are 25% more efficient that will save you 274 kilowatt-hours a year. That’s definitely a good thing, but most people wouldn’t consider the savings from that to be worth the cost of replacing them. It’s possible you could find some other energy efficiency measure would save you more.

      I’m impressed to hear you have an electric car. What kind is it and how do you try to take advantage of your solar output when charging it?

      • Hey Ronald (wow, the man, the myth, the legend!),
        Glad to hear it. 25% is probably close to “replace instead of repair” but not “replace it for fun”.
        Oh mate, don’t get me started on the car!!! I’ve always been a rev-head (still am), but I’m part time stay at home Dad, and my fuel bill was out of control (my 10 year old Falcon didn’t like shopping trips, fuel usage was about 15L/100km!), so my wife suggested I look at electric cars. I was under the impression you could either get a very cool, but VERY expensive toy (cough, Tesla, cough), or the world’s most boring car (Prius). But I’m a good boy, so I did my research, and discovered the BMW i3. And this car is just amazing. Fastest BMW (0-50km/h) meaning I can embarrass just about anything off the line, really comfortable, and a lot more efficient to run vs a Tesla due to the light weight. I get 13kwH/100km, Teslas are 25-30. Managed to get a bargain too, since it was pre-facelift demo, (snobby BMW owners want the new face), had 600kms on it and even talked them into 5 years servicing, 8 years warranty. All with a discount of 35% over a new one!
        Anyway, because I’m at home 4 days a week, it’s easy for me to charge it off the panels (slow charging pulls 2kW, panels are 6kW with 5kW inverter) but I don’t feel guilty if I need to charge it up overnight every now and again.
        Owning it, and letting other people drive it, has already convinced some others in my life to buy electric cars, and the solar panels to go with them. I bought both at (roughly) the same time, they really go together. I’ve told anyone who’ll listen that if you’re a two car family, there’s just no excuse to not get electric. As a shopping cart and daily driver (as opposed to a road tripper or caravan dragger), there really is no comparison.
        Here’s an interesting stat for you: With the amount of electricity we use, the panels will pay for themselves in four years. No surprises there. But, in 8 years time, even if I have to throw this car in the bin instead of selling it, I’ll still have spent less money on it (including purchase price!) than what I would’ve if I kept the Falcon. And that’s assuming the price of petrol (or the greed of those cheeky Saudis) doesn’t go up…!
        Sorry, I’ve rattled on for two long. Let me know if you want a proper review on the i3, and how we use it here in Aus.

    • Mark,

      RCACs will stop working over a certain temperature depending on the gas used. Some as low as 43C. Your manual may note the temp.

      1 Make sure the outside unit is in the shade.
      2 Consider using misters. Coolnsave for an idea of the concept
      3 Cool the space earlier in the day

  21. Hi Finn,
    I have a 2KW system feeding back onto the grid for which I receive a pittance of 11c per kWH. I have been told that I need to run my pool filter during the day as it will run off the solar power generated and not off the grid. How does that work when all my solar generated power simply feeds back onto the grid. Surely the filter will still run off the grid supply at 29.5c per kWH. Can you clarify?

    • Hi there,
      not an expert (nor do I play one in a TV series).

      Assuming you are on a “nett” connection for your solar… A “Nett” connection is where your solar system is set up so that you use your power first, and any excess gets fed back to the grid (at 11c/kWh).

      So, with that established – you use the power in your house (and pool) _before_ it hits the meter and then the grid.

      So, for example, my wife is at home during the day – she uses the washing machine, the dryer (no clothes line), dishwasher and air-cond all through the day and we consume our own “free” (less the cost of the initial install) energy, and any left we export to the grid.

      Our 4kW system means that we pay not a lot during the quarter. Not a “zero bill”, as we still buy on cloudy days and at night, but certainly a lot less.

      Watching the price of battery systems closely, but my spreadsheet still gives me (at best) a 12 year payback for a system that would likely need replacing after 10.

      • Thanks for that, however I don’t think we are on a nett system as our bill shows a rebate for the power we supply back to the grid and the charged for what we use from the grid. If we used our power first then there would be no rebate as our system is a small system and out current daily usage is around 80kWH, which is very high due to the pool and ducted air. What do they call the other system where you simply get paid for what you place back on the grid, and how would we go about changing our system to the nett system? The battery option while intreging, is basically out of the question as we will probably be moving out in 3-4 years.

        • harry wwc says:

          Hi Trevor,

          if you’re not on “Nett” usage, and only getting 11c/kWh you export, then get an electrician out (in NSW a “Level 2 Electrician”) to change the connection to be Nett – i.e. you use your power first, then sell off any excess.

          For me, it cost about $600 to get the meter changed from a ‘standard’ smart meter to a ‘solar nett smart meter’ – and in the 2 1/2 years I almost recouped all that cost. But your mileage may vary depending on what you already have, and what state you are in.

          But either way, you need to make sure you are using your power first, and then selling any excess (if any) to the grid.

          • In addition to what harry says, just bear in mind that your retailer might actually pay for the changeover. In NSW at least the gross to net solar changeover was supposed to coincide with voluntary roll out of smart meters, and handing over responsibility for meters from the distributor to the retailer. If you are with the right retailer, that should be managing that changeover for you and with the right retailer picking up the costs. So check with your retailer. (I was with Red Energy and they did the changeover for me no cost, and I think I lot of other retailers will do the same).

            Part of the reason for this, would be that some people will require a new meter anyway. While changing from gross metering at a high level is just a change of connection from in front of the main meter, to behind it, you also need to make sure that the meter can support import and export, and is configured to do so.

            So definitely contact your retailer and see what they can do for you.

    • Ronald Brakels says:

      Hi Trevor, Ronald here.

      The power your solar produces is used by your home first and then any left over is sent into the grid for the 11 cent feed-in tariff. As you’ll be paying a lot more than 11 cents for grid electricity you save more on your electricity bill when your home uses solar electricity than when it sends it into the grid. So if you run your pool filter during the day you’ll save more on your electricity bills.

      As you have a small solar system and aren’t receiving an old high feed-in tariff that some states still have you may benefit from installing more solar. Let me know if you’d like advice on that.

  22. We have ducted air-conditioning (Acton SRE18C) for which I have been told Cooling power is 5.8kW Heating 5.6kW – this unit is 18 years old, would a more modern unit use less electricity (and how much)? We have it set for 25C in Summer but our generation/consumption graph looks scary when we have those hot days – we have a 6.6kW system.

    • Is this an inverter model? Due to age I am guessing maybe not. In that case you would definitely save on electricity – BUT BEWARE, the replacement cost is not going to be recouped for a long time AND if you have a non inverter model it is MUCH MORE reliable, that’s why it’s already lasted 18 years and will probably last another 18 without touching it. My non inverter has lasted 20 years without problems , YOU DO NOT get that from the new inverter models, there are too many PCB’s, all electronics, you will regret it.

      • Ronald Brakels says:

        Good point. If it’s lasted 18 years then it’s probably a reliable model and may last many more.

    • Ronald Brakels says:


      If you know the “Power Input kW” I can tell you how your air conditioner’s efficiency compares to modern ones. It’s possible you may be able to find it online but usually it’s very difficult to find information on the internet for units that old. But I would expect a modern air conditioner to use around 25% less energy. Possibly 35% less. If you have a problem with leaky ducts the benefits could be even greater. On top of that, if you are the sort of person who likes to use the smart features air conditioners typically come with these days you may be able to improve your solar self consumption and reduce your electricity bills by running it in a low power “eco” mode when it’s not really hot. Whether or not it is worthwhile to replace it will depend on how much you use your conditioner. Many people who only use theirs during summer don’t consider the savings worthwhile to replace an older unit that’s still functioning. But as your air conditioner is so old you may appreciate the reassurance of having a new one with a fresh warranty and being able to replace it at your convenience instead of when it breaks down which will typically be at the worst possible time.

  23. Meredith Lewis says:

    Hi Finn
    I used your website to select a solar PV provider and have been running a 3.8kw system that I’m happy with. But with the power blackout in Melbourne on 25 January due to load-shedding we had 2.5 hours without any power. I’m new to all this and don’t have great analytics on the system (except solar.web, which I find difficult to interpret), but would the PV system continued to have generated power back to the grid during the blackout? And is there any way that the power generated by PV during a blackout could be used in the household without battery storage?

    • Ronald Brakels says:

      Hi Meredith, Ronald here.

      Rooftop solar systems will shut down during a blackout as a safety feature to protect line workers from live cables. This means during a blackout they won’t provide any electricity to the grid. It is possible to pay extra for a hybrid inverter and set it up so it will provide power to a home without batteries during a blackout in the daytime, but this is difficult to do practice for two reasons. Firstly, unless you have a very large solar system it is very easy to try to use too much power causing the whole system to shut down. This can happen because a cloud has covered the sun or the sun is low in the sky and so the system isn’t producing much power. The second reason is when hybrid inverters are used this way they often do a lousy job of it.

      If you want power during blackouts the most cost effective solution is usually to get a small generator. My parents have one for cyclones. It’s not wired into the home so they can lend it to friends or family if they have a problem or take it camping. If they need to they will run an extension cord inside for their fridge and freezer and anything else they may want to power, but they’ve never had an outage long enough to feel the need to use it so far.

  24. Sandra Bastians says:

    Hi…excuse my ignorance (better get that out the way)

    can I split my panels into 2 different sections of the house. ie flat roof (north facing) and west facing gable (adelaide)
    I guess I’m asking if the system I buy can have fronius inverter with panels in 2 different areas so as to take pressure off the flat roof in case they get dirty, being on a flat roof?? or should they all go on the flat(ish) north facing roof?

    Do they act independently if they are in separate areas?
    Also trying not to break the bank. Can all panels achieve this?

    hope that makes some sense 🙂

    • Ronald Brakels says:

      Hi Sandy

      To keep your solar panels clean on a flat roof they should be mounted in tilt frames that hold them at an angle of at least 10 degrees. This article goes into installing on a flat roof:

      It will cost more to install panels on a flat roof due to the extra cost of the tilt frames but they will have higher output then west facing panels. In Adelaide north facing panels at a 10 degree angle will produce an average of 11% more energy than west facing panels at a 20 degree angle.

      If you place some panels facing north and some west they will be able to act independently of each other unless the inverter you are using is very small.

      While you could use any panels I definitely do not recommend getting the absolute cheapest ones. Fortunately, there are low cost panels that are reliable. Our Solar 101 guide has a graphic displaying all the panel brands we consider to be good quality. It has the lower cost brands on the left with the more expensive premium brands on the right:

Speak Your Mind


%d bloggers like this: