Quickly estimating battery savings

The home battery market is newer than the home solar market, and prices are changing rapidly. On top of this, working out savings is more complex. Open Solar fully simulates energy flows hour by hour when estimating battery savings, and is sensitive to a range of factors that can influence the outcome.

This doc steps through a simple approach to estimating battery savings when on a flat tariff, to illustrate the sensitivity of battery savings to these factors.

Open Solar’s model

OpenSolar’s simulation divides the entire year into 8,760 hours and models energy flows in each of these. Separately for each hour, our simulation considers:

• Weather data, including solar radiation for that hour.
• Generation (weather, time-of-day and time-of-year dependent)
• Gross electricity consumption on site for that hour. Even if the provider provides no usage information or coarse-grained info such as a single annual consumption figure, Open Solar estimates an hourly consumption profile using location-specific default data.
• Usable battery capacity versus nameplate capacity
• Flow of energy to and from the battery, taking into account the stored energy at the start of the hour, charge and discharge losses, inverter and clipping losses, and battery capacity degradation over time.
• Import and feed-in price in each hour (should the tariff selected vary over the day or season)
• The tariff assumed for the site prior to getting the battery, and the tariff assumed once the battery is installed, should the user want to simulate a tariff change in parallel with the installation
• Year-to-year changes, like electricity tariff inflation and financial discount rate.

In conducting a detailed simulation like this, we calculate a precise estimate of bill savings and payback. Such detailed calculations are infeasible for a professional to duplicate using rapid back-of-a-napkin” manual calculations.

Nevertheless, a “napkin” calculation is often employed by industry experts for reality-checking simulation results. This is a sensible undertaking, provided the quick estimate adequately considers all factors having a major impact on the outcome. We’ll go through some possible napkin calculations below, and build up accuracy towards the Open Solar estimate to demonstrate the importance of doing this right.

Initial (inaccurate) estimation

The quickest way we’ve seen some professionals estimate battery retrofit savings is to assume a flat import tariff, assume the battery avoids imports equal to its capacity each day, then multiply by 365 days to get annual savings.

Assumptions

• Import price: $0.32 / kWh
• Battery size: 20 kWh
• Fully-installed cost after incentives: $15,000

Annual savings = 20 x 0.32 x 365 = $2,336.

This gives a simple payback period of 15,000 / 2,336 = 6.4 years.

Forgone export income

Without a battery, the excess solar from a site must be sold to the grid. With a battery, this excess solar is not exported, but instead charges the battery. Though feed-in prices are trending downwards and the value of this forgone feed-in is small, it is not zero and must be subtracted from the value of avoided imports:

Additional assumption

• Feed-in price: $0.05 / kWh

Forgone income = 20 x 0.05 x 365 = $365

Annual savings: $2,336 - $365 = $1,971

Simple payback: 15,000 / 1,971 = 7.6 years.

Depth of discharge

Many batteries have a minimum depth of discharge (DoD) to maintain battery health. For example, a 20kWh battery could have a minimum DoD of 10%, giving a usable capacity of 18 kWh. Additionally, some homeowners purchase batteries for blackout protection and set a reserve storage, perhaps at 20%, which would mean 16 kWh of capacity available each day.

Additional assumption

• Minimum depth of discharge: 10%

Annual savings:  18kWh x ($0.32 - $0.05) x 365 = $1,774

Simple payback: 8.5 years

Round-trip efficiency

There are losses when charging a battery, and additional losses when discharging. These include efficiency losses not just within the battery, but also conversion losses in the inverter (and additional conversion losses for AC-coupled systems). Losses while charging mean that it takes more than 18kWh of excess generation to store 18kWh in the battery, meaning that forgone export value is higher. Losses while discharging mean that discharging 18kWh will avoid something less than 18kWh of imports, so the avoided imports value is lower. It’s not necessary to get overly complex here; we can just apply a loss percentage to the overall calculation.

Additional assumption

• Round-trip efficiency: 90%

Annual savings: 18kWh x ($0.32 - $0.05) x 0.9 x 365 = $1,597

Simple payback: 9.4 years

Average cycles per day

So far, we have been assuming that the battery cycles once per day. Except for very small batteries, this is not a realistic assumption. Larger batteries will sometimes not fully charge in a day, and/or full discharge in a day.

Taking Australia as an example: since the introduction of the Australian Cheaper Home Batteries rebate in July 2025, average battery sizes in Australia have risen from a typical size around 10-15 kWh to closer to 20 kWh. These large batteries are hard to fully cycle once per day in standard self-consumption mode, for two separate reasons, as follows.

1. Insufficient excess generation

In Winter and on overcast days, solar generation is lower. Additionally, Winter electricity consumption is often higher than in milder months. Both factors reduce excess generation and increase the likelihood of days when the battery does not fully charge.

In Australia, a household might use 20-30kWh per day on average. It will be less than this in mild months and much more than this in Winter. Even a large-ish 10kW solar power system might only generate around 20kWh per day in Winter. After usage, there is not enough excess to charge the battery. Overcast days at any time of year could see the same result. 

For a site with high Winter usage and modest generation receiving a large battery retrofit, there might be a full quarter of the year where the battery does not fully charge due to insufficient excess generation.

2. Insufficient after-dark consumption

In months where the weather is milder and sunnier, electricity consumption is low, and generation is high, so batteries are likely to fully charge every day. However, to fully cycle, the battery must also fully discharge each day. For a large 20 kWh battery, households may not use that much electricity after dark. A household may fail to fully discharge a large battery for almost half of the year, especially for an energy-conscious household.

Taken together, this means that rather than assuming one full cycle per day, napkin calcs should assume something less. This might be considerably less, perhaps even lower than 0.5 cycles per day for a large battery with modest Winter generation, high Winter usage, and modest after-dark usage in the sunnier half of the year.

Additional assumption

• 0.75 cycles per day

Annual savings: 18kWh x ($0.32 - $0.05) x 0.9 x 365 x 0.75 = $1,197

Simple payback: 12.5 years

Discussion

We can see from the above variations that the payback period difference between the light-weight and inaccurate estimate versus a more detailed estimate is significant. The payback period estimate has almost doubled from 6.4 to 12.5 years. And there are still other factors that can impact the calculations that we haven’t yet considered.

Other factors

Covering these additional factors is perhaps beyond a napkin calculation. Open Solar does consider these aspects in our simulation. Some add to costs, others add to savings:

1. Complex time of use tariffs and demand tariffs
2. Battery maximum charge and discharge power
3. Battery degradation
4. Electricity tariff price inflation
5. Discount rate (i.e., working out payback with real value)
6. More complex battery control schemes, e.g., maximising savings by performing price arbitrage
7. Inverter clipping losses

Recommended napkin calculation

For an estimate that ignores everything too complex and also assumes a simple tariff structure, this formula is a good start:

Annual savings = 
usable_capacity_kWh
x (import_price_per_kWh - feed-in_price_per_kWh) 
x round_trip_efficiency
x 365 days
x avg_cycles_per_day

The only challenging variable to estimate here is the average cycles per day. Here are some suggestions:

• 0.85 or less for modestly-sized batteries
• 0.7 or less for batteries that are large relative to average daily electricity use and generation

You should only consider a higher average cycles per day if:

• The battery is very small (e.g. 5kWh)
• You are confident that electricity consumption is heavily concentrated outside of sun hours.
• Usage tends to be higher in sunnier months than usual (e.g. continuous space cooling in Spring, Summer, and Autumn)
• Electricity consumption is much lower than usual in Winter (non-electric forms of space heating like wood or gas, high thermal-efficiency home, etc)

Additional guidance

As part of its simulation, Open Solar displays simulated average cycles per day in the Battery pane of the Design page, a value that takes into account what the user has entered for all system components and also how they have characterised electricity usage at the site. You can use this in the above napkin calculation if you want to reality-check that calculated annual savings.

If you are looking for further insight into the average cycles per day value we calculate, you can scroll through the indicative daily profile snapshot charts that are available on the Design page in the summary section, and see how battery operation varies throughout the year. For example, if the battery isn’t fully charging in Winter, you will see reduced charging/discharging and more after-dark grid importing.

If the battery is not fully discharged in milder months, there will be no grid importing over the entire night.

No feed-in, minimal battery charge/discharge, and significant evening imports suggest battery not fully charging in Winter.

No grid import means that the battery does not fully discharge after dark in sunnier/milder months.