Round-trip efficiency quietly taxes every unit you store

Defining the number properly

Round-trip efficiency is the energy you get back out of a battery divided by the energy you put in. A figure of ninety per cent means that storing ten units returns nine, and the missing unit has become heat somewhere in the chain.

The number is straightforward to state and surprisingly slippery to compare, because manufacturers measure it at different boundaries. A cell-level figure counts only what happens inside the cells and looks excellent. A system-level figure measured from the household wiring, through the inverter, into the cells and back out through the inverter again, counts everything and looks less excellent. Both are honest and they are not comparable.

When a quote gives an efficiency figure, the useful question is where the measurement started and stopped. If nobody can tell you, treat the number as marketing.

Where the losses actually occur

The cells themselves are the good part of the chain. Lithium cells have low internal resistance and lose only a modest fraction of what passes through them, and that loss rises with current, which is one more reason a battery pushed hard performs slightly worse than one working gently.

Conversion is where most of the shortfall lives. In an AC-coupled system, solar output is converted from DC to AC by one inverter, converted back to DC to charge the battery, then converted to AC again to run the house. Three conversions, each taking its percentage, and the percentages multiply rather than add. A DC-coupled system with a hybrid inverter skips one of those steps by charging the battery directly from the array’s direct current, which is a real and measurable advantage.

Then there is everything that is not conversion at all: the management electronics, the communications, the cooling fans and, in cold climates, the heaters that bring cells to a temperature where they will accept charge. None of that appears in a cell-level efficiency figure and all of it comes out of your meter.

Idle consumption matters more than the headline figure

For a system that cycles fully every day, conversion losses dominate and the round-trip percentage is the number to watch. For a system that cycles rarely, the standing consumption dominates completely, and a good round-trip efficiency is close to irrelevant.

The arithmetic is easy to sketch even without a price. Suppose the electronics draw a small but continuous load. Over a day that is a fixed quantity of energy, regardless of whether the battery moved anything. If the battery shifted a substantial amount that day, the fixed cost is spread thinly. If it shifted almost nothing, the fixed cost stands alone against nothing, and the system consumed more than it saved.

That is not a hypothetical failure mode. A large battery installed on a small household load, or one held mostly in reserve for outages, can plausibly be a net consumer of energy on quiet days. It may still be worth having for the resilience, but that is a different justification and it should be stated as one.

The tax sets a floor on when storing is worthwhile

Storage makes economic sense only when a unit is worth meaningfully more at the time it comes out than at the time it went in. Round-trip efficiency raises the bar for that comparison, because you must put in more than you take out.

Take an illustrative case with no prices in it. If the round trip returns nine units for every ten stored, then every unit delivered at the expensive time cost you slightly more than one unit at the cheap time. The gap between the two prices must therefore cover that shortfall before it covers anything else, and only what remains after that is available to repay the hardware.

Where import and export prices are close together, that gap may be too small to clear even the losses, and storing solar output rather than exporting it can genuinely lose money. Where a time-of-use tariff creates a wide spread between cheap overnight and expensive evening rates, the gap clears the losses comfortably and the calculation looks entirely different. This is jurisdiction-specific and it changes when tariffs change, which they do.

Reading a real system honestly

If you already have a battery and a monitoring system, the round-trip efficiency you are actually achieving is measurable. Compare the energy recorded going into the battery over a month with the energy recorded coming out, and include whatever the system reports as its own consumption if it reports it at all.

The result will almost certainly be worse than the datasheet, and that is not a defect. The datasheet figure is measured at a favourable state of charge, at a favourable temperature and at a favourable power level, and your house does not operate there.

What the measurement is good for is deciding whether the battery is doing enough work to justify its standing costs. A month where throughput was low and idle consumption was proportionally high is a month the battery cost you something. Several of those in a row is evidence about sizing, and it is better to have that evidence than an assumption.