Cycle Life & Round-Trip Efficiency

Round-trip efficiency in plain terms

Think of a battery as a bucket with a small leak. Every time energy is poured in and poured back out, a little spills. Round-trip efficiency is the share that makes the trip. A 90% efficient battery returns 9 kWh for every 10 kWh used to charge it; an 83% efficient one returns 8.3 kWh.

Three things cause the spill. Converting household alternating current (AC) into direct current (DC) for the cells and back again costs a few percent each way. The cells themselves warm up as current flows. And the battery’s electronics, cooling and communications draw a small amount of standby power all day, even when the battery is idle. That last item explains much of the gap between lab ratings and field measurements.

Rated vs. measured efficiency

Manufacturers measure efficiency differently, so ratings are not directly comparable. Tesla lists 89% for Powerwall 3 from solar to battery to home, and 97.5% for solar going straight to the home, which skips the battery (Tesla Powerwall 3 datasheet). FranklinWH rates the aPower 2 at 90% from grid to battery to load at 3 kW and 25°C (FranklinWH aPower 2 datasheet). Enphase lists 90% AC and 96% DC for the IQ Battery 5P (Enphase IQ Battery 5P datasheet). SolarEdge lists a 94.5% peak for its DC-coupled Home Battery, which shares an inverter with the solar array (SolarEdge Home Battery 400V datasheet). Generac lists 88% for PWRcell 2 (Generac PWRcell 2 specification sheet), and sonnen lists 85.8% using California’s SGIP formula (sonnenCore+ technical specifications). LG does not publish a round-trip figure for the enblock S (LG enblock S datasheet).

Independent measurements come in lower. California’s Self-Generation Incentive Program (SGIP) evaluation found an 83% average and 82% median across 436 residential systems that cycled about 117 times a year (CPUC SGIP 2019 storage impact evaluation); the 2020 evaluation again found 83% (CPUC SGIP 2020 storage impact evaluation). An earlier evaluation measured about 40 W of average standby draw (CPUC SGIP 2017 storage impact evaluation). Enphase’s own white paper estimates 84–86% real-world efficiency for the 5P against its 90% rating (Enphase battery efficiency white paper). A HTW Berlin lab study of 26 systems measured standby draw from 2 W to over 70 W, with a median of 12 W (HTW Berlin study of 26 home storage systems).

The economic test: the break-even ratio

Shifting energy pays only when the kWh displaced is worth more than the kWh used to charge, after losses. The rule is simple: the off-peak price divided by the efficiency must be lower than the peak price. Put another way, the break-even efficiency is the off-peak price divided by the peak price. Below it, every cycle loses money before wear is counted.

On SCE’s TOU-D-PRIME plan, summer weekday peak energy is 61¢ and off-peak is 26¢ (SCE TOU-D-PRIME plan), so a battery breaks even at 43% efficiency and almost any battery clears it. On a tariff with a narrow gap, such as PG&E’s E-TOU-C or DTE’s Time of Day plan, the break-even figure sits above 70% and field efficiency of 83% leaves only a sliver of value.

The dollar cost of losses scales with how much a battery cycles and what charging energy costs. A battery delivering 10 kWh a day loses about 210 kWh a year at 94.5% efficiency and about 750 kWh at 83%. At a 30¢ charging price, that is the difference between about $64 and $224 a year.

Cycle life and warranted throughput

A full cycle is charging a battery from empty to full and back once. Each cycle wears the cells slightly, so warranties limit use by years, by cycles or by total energy delivered, called throughput. Dividing a throughput cap by capacity converts it into full-cycle equivalents for comparison.

FranklinWH warrants the aPower 2 for 15 years or 60 MWh (FranklinWH aPower 2 datasheet). Enphase warrants the 5P for 15 years or 6,000 cycles, with more than 60% of capacity remaining (Enphase IQ Battery 5P datasheet). sonnen warrants the sonnenCore+ for 10 years or 10,000 cycles (sonnenCore+ technical specifications). Generac warrants each 3 kWh PWRcell 2 module for 10 years or 7.56 MWh (Generac PWRcell 2 specification sheet). SolarEdge warrants 70% capacity at 10 years with no cycle limit (SolarEdge Home Battery 400V datasheet). Tesla guarantees 70% capacity at 10 years with unlimited cycles for self-consumption, time-of-use and backup, but caps any other use, which includes VPP participation, at 37.8 MWh of throughput (Tesla Powerwall U.S. warranty).

Wear cost: the second economic test

Dividing installed cost by warranted throughput gives a rough wear cost for each kWh delivered. Using EnergySage’s brand averages of $1,232 per kWh for FranklinWH, $967 for Tesla and $1,429 for Enphase (EnergySage battery cost data), the warranty-based wear cost is about 24–35¢ per kWh. That is a ceiling, not a forecast: many batteries outlast their warranty terms, and a battery bought for backup carries its cost regardless of cycling.

The comparison still clarifies who should cycle a battery hard. A program or tariff that pays less per kWh than the wear cost is converting warranty life into revenue at a loss, unless the battery would otherwise sit idle until its calendar warranty expires. This is one reason VPP programs usually pay per kW of capacity for a limited number of events, rather than per kWh of daily cycling.