Wind turbines, solar panels, home battery storage - if it's discussion about renewable energy you're after, you'll find it here.
Updated to 31st July 2018
Please read An Introduction to AC and DC Coupled Solar Battery Systems for an understanding of terms used in this document.
In this article I have compared 3.50 years of data from 5.2 kW of REC solar panels and an SMA 5000TL-21 solar inverter, installed on November 17th 2013, against 1.17 years of the same panels with an 11.6 kWh Sunverge SIS-7048 DC coupled solar battery, installed on 26th May 2017 as a part of the AGL Virtual Power Plant (VPP).
As a retired electrical engineer, I have always assumed that adding solar battery storage would result in significant savings, so I was amazed to find the addition of the battery is actually costing us money, no matter which way the data is examined.
The critical point to remember in solar battery economics is that the cost of a battery can only be covered by a reduction in imported peak energy costs, or an increase in exported feed-in revenue, or ideally both. It is also very important to keep in mind that this is a case study based on comparing solar panels only to solar panels plus a battery. It is not a study based on a new installation of panels and battery.
This is a detailed comparison of energy flows and costs for the period without a battery, versus the period with a battery, and provides three very clear reasons as to why we didn’t save money, and are not likely to do so in the future.
ENERGY FLOW RATIOS: The most significant, and surprising, reason why adding a battery did not save any money turns out to be based on the way our imported and exported energy flows changed with the addition of the battery, as Table 1 below shows.
Row 1 shows the current tariffs, Row 2 shows the annualised peak imported and feed-in exported energies from December 2013 to May 2017 for panels only, and Row 3 shows the annualised peak imported and feed-in exported energies from June 2017 onwards for panels and battery.
Row 4 shows the ratio of how much annualised imports and exports have changed as a result of adding a battery, and Row 5 shows the annualised reductions in kWh.
Applying the tariffs in Row 1, to the energy changes in Row 5, results in the annualised cash flows in Row 6.
Row 7 is Row 6 export revenue changes minus Row 6 import cost changes, and shows how much per annum we are worse off as a result of adding a battery.
Row 6 clearly shows that we have lost more in export revenue than we have gained in reduced import costs, as reflected in the ratio of 1.368 . If we are to pay off the battery at all, we need the ratio in Row 6 to be less than 1.000. That is, we need to save more in imported energy costs than we lose in exported revenue.
Table 1 shows that we export roughly 3 times more than we import, regardless of whether we have a battery or not, and that adding a battery roughly halves both imports and exports. These symmetries suggest that this is inherent in the physics of adding a battery, and if so, there is very little we can do about that. The logical conclusion then is that it is the tariffs that will determine whether we can pay off our battery, or not.
A second way to illustrate the problem is to assume ratios of 3 and 0.5, and start with 6 units of exported energy and 2 units of imported energy for panels only. Adding a battery halves those numbers to 3 units of exported energy and 1 unit of imported energy. In other words, at AGL’s current discounted tariffs of $0.16300 for exports and $0.35759 for imports we have gone from a profit of 6 x $0.16300 – 3 X $0.35759 = $0.26282 with panels only to a profit of 3 x $0.16300 – 1 X $0.35759 = $0.13141 with a battery. Adding a battery has halved our cash flow, and that means there is no way to pay off the battery.
We can apply some simple mathematics to try and work out what the tariffs would need to be in order for us to start paying off the battery. The change in cash flows ratio in Row 6 links the tariffs ratio in Row 1 and the changes in energy flows ratio in Row 5 in that 1.368 = 3.002 X 0.456. The break-even point is when the ratio in Row 6 becomes 1.000. Dividing both sides of the equation by 1.368 gives us 1.000 = 3.002 X (0.456 ÷ 1.368 ), or 1.000 = 3.002 X 0.333. This means that 0.333 is the ratio of export tariff to import tariff at which we would gain as much from reduced import costs as we would lose in export revenue.
If the net import tariff were to remain at $0.35759, the export tariff would need to be less than $0.35759 X 0.333 = $0.11911 before we can begin to pay off our battery.
If the net export tariff were to remain at $0.16300, the import tariff would need to be more than $0.16300 ÷ 0.333 = $0.48936 before we can begin to pay off our battery.
The problem with both of the above required tariff changes is that the direction they need to go is counter to current trends. The government is trying to bring down peak import prices, and to encourage solar panels by increasing the export tariff. Going forward therefore, it is more likely that the loss of revenue from adding a battery will increase. As an example, AGL reduced the peak tariff on July 1st 2018 from $0.35948 to $0.35759; not a big reduction, but going the wrong way for us in terms of battery economics.
Another way to see if we have had any economic benefit from adding a battery is to look at the two periods on the basis of what we actually paid, and also look at what would have happened if we adjusted past tariffs to reflect current tariffs. Both of these methods also showed losses, as indicated in Table 2 below.
Columns 1 to 4 summarise what we actually paid, and what the average tariffs were, according to our bills.
The large negative annual benefit shown for Columns 1 and 2 is a result of much lower import tariffs and much higher export tariffs for the period when we had panels only, compared to the period when we had the battery as well.
Columns 7 and 8 show the percentage changes in average tariffs for both periods, and when the costs in Columns 1 and 2 are scaled by these changes, we get the costs in Columns 9 and 10. Note that the tariff-adjusted negative annual benefit for Columns 9 and 10 is very similar to that of Row 7 in Table 1, as would be expected.
An important distinction to make in this analysis is that we still save some money, whether or not we have a battery, as Column 11 shows (Column 10 – Column 9). The point is that we save a lot less by adding a battery. The solar panels are the real driver of solar savings, as I show in The Significant Cost Benefits of Solar Panels.
Also, the difference between actual costs and current costs in Table 2 highlights the difficulty of a simple price comparison over long time frames as a method of measuring any cost benefit from adding a battery.
SOLAR ENERGY STORAGE LOSSES: Another big surprise was just how much solar energy is wasted by the process of saving energy in a battery, and reusing it later via a battery inverter. These losses are a major part of why we didn’t generate enough feed-in revenue, as per Tables 1 and 2 above.
The round trip losses for the period for which we had only solar panels consisted of the SMA solar inverter losses of 3.5%, as specified by SMA, since they can’t be directly measured.
The round trip losses for the Sunverge battery system were measured at 22.7%. Adding 4.0% for the specified losses of the Sunverge Conext MPPT 80 600 solar charge controller makes a total of 26.7%.
With no battery, we generated a total of 8,869.5 kWh per annum, of which 299.9 kWh was lost.
With a battery, we generated a total of 8,464.3 kWh per annum, of which 2,263.0 kWh was lost.
This is a 654.5% increase in losses.
At the current feed-in tariff of $0.16300, these losses equate to $48.89 and $368.87.
SUNVERGE BATTERY MANAGEMENT SOFTWARE: A detailed examination of the way the Sunverge battery software managed the importation of peak energy showed that we imported 59.6% more peak energy than we required in the 426 days from 1st June 2017, at an extra cost of $303.07 at the then prevailing AGL discounted tariffs, or $312.10 at the current discounted peak tariff. That is, of the total of 1,465.2 kWh of imported peak energy, 872.8 kWh was excess to requirements.
The details of how these numbers were derived are contained in the article VPP Solar Battery Management Software Imports 60% More Than Required.
We installed 5.2 kW of 260 watt REC solar panels on November 17th 2013, as detailed in The Significant Cost Benefits of Solar Panels. The panels face directly north, they are not shaded, and they are at a pitch of 22.5 degrees above the horizontal. We also had installed an SMA 5000TL-21 solar inverter from which I downloaded daily solar production data via Bluetooth.
We added a Sunverge SIS-7048 DC coupled solar battery when we joined the AGL Virtual Power Plant (VPP) on 26th May 2017. Solar production and consumption data were then taken from AGL’s Solar Command platform.
As a retired electrical engineer, I have maintained a very detailed spreadsheet of all our monthly electricity costs and energy flows since November 17th 2013, with the earlier quarterly bills converted on a pro rata basis into monthly data. The analyses in this document are based on two retired people living in a three bedroom home, and all prices include GST, unless otherwise specified.
All spreadsheet data has been reconciled to the cent and to the watt, and are based entirely on actual data except for January 2018, when a Sunverge software update failed and the system was disabled for 10 days. This meant we imported a lot more peak energy than normal, and exported a lot less feed-in. For January therefore, I have used the daily average of the 21 days for which we had solar energy scaled up to 31 days.
The gross peak tariffs quoted in this document are the full retail tariff including GST. The net peak tariffs are calculated by first discounting the ex-GST rate at the offered discount rate, and then adding 10% GST. Government feed-in tariffs are always ex-GST.
I have excluded state government pensioner discounts and other minor account-based charges, and I have omitted controlled load in this document because it is not pertinent to solar energy generation and storage.
This analysis only applies to the situation where a battery has been added to an existing solar panel installation. If panels and a battery are installed at the same time, you would expect to see the benefits outlined in The Significant Cost Benefits of Solar Panels, but the cost benefits of the battery would be impossible to isolate.
ENERGY FLOW RATIOS
Table 3 below shows the costs, exactly as they occurred, based on having solar panels installed, but no solar battery. The PER ANNUM data at the bottom of Table 3 is derived by dividing the TOTAL values by 1278 to give a daily average, and then multiplying by 365 to give an annualised average.
Table 4 below shows the costs, exactly as they occurred, based on having solar panels and a Sunverge solar battery. The PER ANNUM data at the bottom of Table 4 is also derived by dividing the TOTAL values by 426 to give a daily average, and then multiplying by 365 to give an annualised average.
Tables 1 and 2 above show just how much the change in energy flows affects the economic viability of solar battery storage. In order to understand how applicable these results might be to other installations, we need to look at the variables involved. I think it is reasonable to assume that most households will not change the way they use electricity, regardless of whether they have panels only, or panels and a battery.
Table 5 above shows our annualised total consumption of electricity, not including controlled load, but including solar and imported peak grid energy. The data for 2018 is for a partial year. Column 6 is the solar energy used in-house (Column 4) as a percentage of the total energy used in-house (Column 5). This indicates an average of about 40% for the period when we had panels only, and has increased to around 60% to 80% with the addition of a battery. For the data in Column 4, it is not possible to isolate the solar energy actually used in-house from the solar energy stored in the battery.
Clearly, a lot of the solar energy is now going into the battery, instead of being exported. However, remember that by adding a battery, we have lost a further 23.2% of all our solar energy through losses.
The main point in Table 5 is that our average annual household consumption is 4,224 kWh each year.
The following table suggests the possible effect of some variables, not in any order of significance, and how they might alter the ratios in Table 1.
Tariffs will have no effect on energy flow ratios, but they will determine how those ratios affect final costs, as highlighted in the Table 1 analysis.
AC or DC coupled battery
The process of converting light to electricity, storing it, and reusing it, is essentially the same for both, so I don’t think this would make a significant difference. There will be some very small difference in losses, but I don’t believe that it will be very significant.
A larger battery would mean that more solar energy will be stored, thereby reducing imported energy costs, but also reducing exported energy revenues. I see no reason to assume this will change the overall ratios in Table 1. In fact, if you consider just panels as a zero battery size, and you look at the symmetries in Table 1, it suggests the main characteristics of a battery are added annual costs, a 654.5% increase in losses, and a small reserve for lights and a fridge in the event of a blackout.
Everyone in the VPP presumably has the same 8.1 kWh battery size for the period of this study. I will be replacing the 8.1 kWh Sunverge battery with the Tesla Powerwall 2 battery, with its guaranteed 13.5 kWh useable capacity, so we will be able to review the effect of battery size over the next year or so.
Solar panel array size
A smaller array would mean more imported energy and less exported energy, with a lower initial capital outlay to recoup, but it would require detailed case studies on different sized arrays to draw any conclusions. However, all those in the VPP, as I understand it, have been assessed as having sufficient capacity to warrant a battery, and I imagine a lot would have 5 kW of capacity. I would recommend that, where possible, and given current prices, most people should be installing the SA Power Networks single phase legal limit of 5 kW, regardless of whether they intend to add storage.
Annual household consumption
Increasing household consumption may cause the battery to be drained more, resulting in reduced peak import costs, but that means feed-in export revenues will be reduced the next day in order to recharge the battery. The greater use of the battery will also increase losses due to charging, discharging and inverting. I don’t see any clear mechanism whereby this could have a significant impact. If the ratios in Table 1 are inherent in adding a battery, I don’t think annual consumption is likely to be a big factor, but it clearly is an unknown variable.
Time of use of solar
This might have some effect, but any daytime solar used in-house will reduce imported energy, and reduce exported energy as well, so I wouldn’t expect it to significantly alter the ratios in Table 1.
SOLAR ENERGY STORAGE LOSSES
I should point out that this analysis is for a DC coupled Sunverge system, so I am not able to say precisely how this would apply to an AC coupled Sunverge system. However, both Sunverge systems have a transformer-based inverter with a very heavy iron core transformer (the inverter alone weighs over 70 kg) which is very inductive and relatively inefficient, and also results in a lot of heat being generated.
Most modern inverters use electronic conversion without transformers and are therefore likely to be more efficient. I will be updating this analysis on a monthly basis so that we can see how the Fronius Primo solar inverter and the Tesla Powerwall 2 compare with the Sunverge system. See An Engineer's Choice for the Next Stage of the VPP.
Table 6 above shows how the round trip efficiency was calculated.
The data in Columns 2 and 3 were taken directly from our AGL monthly bills. Columns 4 and 5 were downloaded in Excel csv format from AGL’s Solar Command. Column 6 is the total solar energy available for in-house use, after allowing for the energy exported. Column 7 is the total solar energy used in-house, after allowing for peak energy imported. Column 8 is the difference between the total solar energy available and the solar energy actually used. Column 9 shows the losses in Column 8, expressed as a percentage of production in Column 4, and Column 10 is 100% minus the losses, or the round trip efficiency.
While the dataset is limited, Column 9 suggests that losses may be significantly higher in winter, by about 8%.
The losses in the MPPT charge controller are not captured in this analysis because the measurement of solar production is done at the battery, downstream from the panels and the charge controller. The Sunverge Conext MPPT 80 600 solar charge controller specification sheet states an efficiency at 48 volts of 96%, which equates to an additional loss of 4%.
The final round trip losses are therefore 26.7%, which means 2,263.0 kWh of wasted energy out of a total of 8,464.3 kWh. This equates to $368.87 at the current feed-in tariff of $0.16300.
SUNVERGE BATTERY MANAGEMENT SOFTWARE
The details are contained in the article VPP Solar Battery Management Software Imports 60% More Than Required.
I had never thought about energy flow ratios before, but they became blindingly obvious once I looked at the data. It is quite possible that most self-consumption solar battery systems are just not economic at current tariffs. I am hoping to receive feedback from participants, so we can build a more comprehensive understanding of the economics of solar batteries, and how variable the ratios might be.
I have been very surprised at how much energy is wasted by the process of charging, discharging, and reusing solar battery energy. The economic losses following on from these energy losses are very significant, particularly in light of the energy flow losses highlighted above.
There are two more aspects of solar batteries that I haven’t taken into account.
The first is their limited lifetime and replacement costs. The REC solar panels we have used have a 10 year warranty and a 25 year linearity warranty. This means they are warranted not to degrade by more than 0.7% per annum, which means that after 25 years they will still perform at 84% of their original specification. Most solar batteries seem to have a 10 year warranty, but often with no linearity warranty. The Tesla Powerwall 2 does, at least, warrant 70% of the initial 13.5 kWh will be available after 10 years.
The second is that replacement costs and repair costs for solar battery systems will clearly be much greater than for a simple solar panel array without a battery. We have had two installation faults and at least three run-time failures so far with our Sunverge battery, so if we were responsible for the costs, that could be a very significant additional cost.
One of the hugely annoying things about the Sunverge SIS for me has been the noise levels from the fans. Again, I am talking about the DC coupled SIS-7048, but I can imagine the AC coupled system is similar. I’ve measured the noise at 20 mm from the inlet and outlet fan vents on the Sunverge and they run at an average of 72.3 dB on the inlet and 75.5 dB on the outlet. That is equivalent to being inside a car on my Decibel X PRO app. Incidentally, I’ve measured the temperature on a hot day at the top of the inverter and it was 58 degrees with fans running. This means the inverter will have derated itself significantly, so I’m not able to generate and export as much as I could from a cooler running system. I believe the massive iron-cored transformer in the Sunverge hybrid inverter is a significant source of the losses observed, and a contributor to the excessive heat and noise levels.
In all the time we have had the Sunverge system running, I have never seen the output of the inverter exceed 4.6 kW. We have 5.2 kW of panels on the roof, and while 5.2 kW is never likely to be achieved because of negative temperature coefficients in the panels, panel ageing and other factors, we had seen power levels over 5.0 kW before Sunverge, when we had the SMA solar inverter. The Sunverge claim that it is a 5 kW inverter under normal operating conditions appears to be incorrect. This means we have been missing out, albeit at times only, on up to 400 watts of solar power.
Unfortunately, just comparing past bills will not really give you a precise understanding of how much you may have lost or gained by installing a battery. This is most evident from the differences between actual costs and current costs as shown above.
I am very keen to see how widely applicable this case study is, so if you are interested in helping, you can create your own table, and send me the results:
Take care with the process of annualising to make sure you have correctly added the days and energies for panels only, and for panels plus battery. There is no need to take account of historical tariffs, because, as I have shown, it all comes down to the ratios and current tariffs.
It seemed such a good idea to install a solar battery, and it felt good to do so, but unfortunately, so far, it hasn’t been worth it. Imagine paying the full price of $26,000 for the Sunverge battery, as in battery comparison table, and finding out that it has cost you money to run it. Even at the heavily subsidised VPP price, it appears to be a questionable investment. I had no idea of the extent of the problems identified in this article when I wrote A Cost Benefit Analysis of the VPP Using 3 Years Realtime Data and Why I, as a retired electrical engineer, joined the AGL VPP.
I will be updating this document regularly so it will be interesting to see if the Tesla Powerwall 2 (An Engineer's Choice for the Next Stage of the VPP) is any better on efficiency and performance. Also, my Fronius Primo inverter with Smart Meter will give me revenue grade data on imported and exported energy, so I will be able to drill down into the time of day, to see when any excess peak energy is being imported, rather than wait in the forlorn hope that AGL will one day upgrade Solar Command.
If, after reading this article, you are still thinking of not installing solar panels until battery technology becomes much cheaper, I strongly recommend you to read The Significant Cost Benefits of Solar Panels.
Another form of solar battery worth consideration is solar hot water, as in The Significant Cost Benefits of Solar Hot Water.
Please see my Profile if you’d like to learn more about other articles I have written on solar energy and my participation in AGL’s Virtual Power Plant solar battery storage program.