Let’s face it, most customers who are buying batteries at the moment are early adopters that aren’t particularly concerned about return on investment (ROI). After all, if they were seeking a high ROI, they wouldn’t be interested in the 10-plus year payback that is common on most batteries.

Common purchasing motivations include reducing solar export, minimising consumption during peak tariffs and reducing their dependence on the grid. But the bottom line is they are most certainly interested in reducing their electricity bill.

The key question for those people selling batteries is: “How can I configure batteries to maximise savings on my customer’s electricity bill, without adversely affecting the battery lifetime?”

The hidden problem is that many of the strategies that sound like common-sense may actually set customers back. For example, charging your battery from off-peak tariffs so you can then discharge during peak tariffs may mean your battery can’t soak up as much excess solar power. Even if you’re able to cycle your battery multiple times per day, this can impact upon your battery life.

In this article we will run through the most common strategies in battery control to identify which produces the best outcomes for an end-user, using SunWiz’s PVsell software. We find:

  • Smaller is better: smaller batteries pay for themselves sooner, so have a higher ROI.
  • In general it’s best to discharge during peak (and shoulder, where applicable) but there isn’t much additional benefit from discharging during off-peak periods.
  • Charging from off-peak tariffs typically reduces the soak-up of excess solar energy to an equivalent amount, and shortens battery lifetime while increasing grid dependence.

Smaller is better

While batteries remain so expensive, the ROI from a PV-only system will always be better than that of a PV-storage system. After all, PV is generating energy whereas batteries are just shifting it. So, if a $7,500 PV system has a five-year payback, adding a battery that costs more than the PV system and produces less financial benefit is going to dampen the ROI. A bigger battery that costs more will harm the ROI more than a small battery with lower price, as it’s the PV that’s doing the heavy lifting when it comes to ROI, and the battery which is acting as a drag.

The second reason a PV system with a small battery has a better ROI than one with a large battery is because battery ROI closely corresponds with battery utilisation. If part of your battery ever sits idle for a day, then it’s not generating revenue, which means it’s not paying itself back. On cloudy days the battery won’t fully charge, and on days of low overnight consumption the battery may not fully discharge. This means the first kWh in the battery is the more utilised than the second kWh, which is more utilised than the third kWh, etc. Hence smaller batteries are more highly utilised, and therefore generate greater income per dollar invested.

The chart below shows the relationship between ROI and annual benefits (bill reduction) to the battery capacity. Here we’re using a residential customer with high consumption (25kWh/day) much of which occurs during the daylight hours, and an Origin electricity bill in the Ausgrid (NSW) distribution network. The system price is $7,090 (including GST) for a 5kW PV system, plus $1,100/kWh of storage. You can see that the ROI is greatest (18%) when there is no battery, and it diminishes to 13% with 6kWh of storage.

The upshot is that it makes fcapture-1inancial sense to purchase smaller batteries. Witness LG Chem’s new 3.3kWh battery; and Enphase is having plenty of success from a micro-battery that can be progressively installed in small increments or perfectly tailored to a household’s energy consumption level.

Discharge during peak, not off-peak

We’ve seen that a highly-utilised battery is going to pay for itself quicker than a battery with low utilisation. However, that doesn’t mean it makes sense to discharge the battery at any opportunity. Because cycling the battery will cause some wear and tear, we only want to utilise the battery when it’s profitable to do so.

Solar power is the primary means of charging batteries — energy that would otherwise be exported and earn a 5-8c/kWh feed-in tariff. Clearly discharging the battery during peak tariffs when you can earn 28-52c/kWh makes financial sense. But does it also make sense to discharge during off-peak periods, when you can earn 12-16c/kWh?

Discharging at off-peak rates empties the battery so it is available to soak up more solar power the next day, so in theory you’re increasing the value of solar energy from the FiT rate to the off-peak rate (that is, from 5-8c/kWh to 12-16c/kWh). But you will be substantially increasing battery wear and tear by shifting a lot of energy through it and discharging it completely for only a small financial gain.

This will accelerate battery degradation, which will mean that it doesn’t perform as well next year and the year after that. So while you may gain an extra 4-11c/kWh from discharging at off-peak rates, your battery’s capacity to earn peak rates will be diminished for years to come.

An example will illustrate this – using a 6kWh battery and the same parameters from the previous example, when we opt to discharge during off-peak periods as well, our annual savings reduces from $1,812 to $1,800 but we increase battery throughput by 4%.


Don’t charge from off-peak

On face value, charging the battery at off-peak rates of 12-16c/kWh and then discharging during peak tariffs of 28-52c/kWh also sounds like a sensible strategy, to the point it has a name of its own: “arbitrage”. Many battery companies are selling the benefits of multiple cycles per day. However, for most residential applications, arbitrage conflicts with the goal of soaking up excess solar power, and actually reduces grid independence.

Pre-charging your battery overnight at off-peak rates, then discharging as soon as peak tariffs kick in is a strategy that would work, if it weren’t for the fact that most residential PV systems are already exporting energy by 8am. Our analysis of thousands of interval-metered homes shows that most households average 1-2kW of consumption between 6am-9am. You may be able to fully discharge a 6kWh battery in those three hours if your solar system wasn’t generating power, but your typical 5kW solar system is already offsetting load by 8am and exporting power by 9am.

Most households can’t discharge a battery in the morning before they start re-charging the battery from excess solar energy. If you start the morning with a battery already full of energy you paid 12-16c/kWh for, then you’ll export all your excess solar power (at 5-8c/kWh) and make a financial loss.

Not only will you have made a loss by vainly attempting to cycle your batteries multiple times per day, you will have increased your dependence upon the grid (by importing more power to charge the batteries), and you may have shortened your batteries’ life by increasing the gross amount of energy throughput over the course of a year.

Again using our base case as an example (6kWh of storage, discharge during peak and shoulder), when we pre-charge the batteries from off-peak power to 50% of their capacity, these charts illustrate that arbitrage has reduced the ability of the battery to soak up excess solar energy. In this example annual savings reduce from $1,812 to $1,807 though we increase battery wear and tear by 20% and increase grid dependence by 12%



These results vary considerably depending upon individual circumstances: consumption patterns, panel orientation, electricity prices and tariff structure, and the intelligence of your battery’s inverter. And so far we’ve focussed or intention primarily upon maximising ROI, or equivalently “How can I configure batteries to maximise savings on my customer’s electricity bill, without adversely affecting the battery lifetime?”

But early adopters are less focused on ROI and more focused on increasing grid independence and minimising their electricity bill. This sometimes means doing things that may be sub-optimal from an ROI perspective, such as discharging during off-peak periods and charging from off-peak. To accommodate their desires, you’ll want an inverter with intelligent control settings, as we’ll see.

Many time-of-use tariffs have off-peak pricing over the weekend. This means the battery will sit idle for two days a week. While there is little financial gain to be made from discharging at off-peak prices, if you allow the battery to discharge at off-peak prices then at the very least you’ll be improving the customer’s grid-independence. And if you have an inverter smart enough to do this only over the weekend, you may strike a good balance between achieving grid-independence and battery wear-and-tear.

Though most households will be worse off from performing arbitrage, there are some exceptions.

  1. Firstly, businesses that have a high morning load will have hours before the solar power system begins to export, providing a good opportunity to perform arbitrage before the batteries are needed to soak up excess solar in the middle of the day.
  2. The same principle applies for households with very high morning load, or more precisely any PV system that is appropriately sized to the load such that it only exports power in the middle of the day.
  3. There will always be cloudy days where your PV system doesn’t contribute much to household load, let alone export any power. The battery will otherwise sit idle on these days, making cloudy days a great opportunity to perform arbitrage. If your inverter can access a weather forecast, then it could intelligently perform arbitrage only on days where it is profitable to do so. And two arbitrage cycles may be profitable if there is a large price differential between peak and shoulder electricity prices and between shoulder and off-peak prices, in which case you could charge the battery from shoulder tariffs on days when the sun doesn’t fully charge the battery.

The key take-home message is that every common-sense strategy must be simulated and tested, as flow-on effects can quickly negate the benefit you were attempting to capture. Because the results vary so widely depending upon your customer’s specific circumstances, it is necessary to customer-specific simulations.

Warwick Johnston is managing director of SunWiz, which owns the cloud-based optimisation tool PVsell (www.pvsell.com.au)