Round-Trip Efficiency: The Energy You Get Back

Round-trip efficiency is the least glamorous number on a battery datasheet and quietly the most expensive. It answers one blunt question: of the energy you pushed into the system, how much do you get back out? A cabinet that returns 95% and one that returns 88% look almost identical in a brochure — seven points is easy to shrug off. Over a decade of daily cycling, though, that gap is thousands of full charges of electricity you paid for and never sold. This guide walks through where the energy actually goes, why the way the number is measured changes it, and how a couple of percentage points compound into real money.

The short definition

Round-trip efficiency (RTE) is energy out divided by energy in over one complete charge-and-discharge cycle, expressed as a percentage. Put 100 kWh into a system, get 93 kWh back, and its round-trip efficiency is 93%. The missing 7 kWh did not vanish — it turned into heat, ran a cooling fan, or was consumed by the electronics that make the system work. Every one of those losses is a real cost, and unlike the sticker price you pay for them again on every single cycle for the life of the asset.

RTE is a whole-system property, not a cell property. A bare LFP cell is electrochemically very efficient. By the time that cell sits inside a cabinet with an inverter, a cooling loop and a control system wrapped around it, the number you actually measure at the terminals is several points lower. Which terminals you measure at is the whole game — more on that below.

Where the energy actually goes

Round-trip losses come from four places, and knowing which one dominates tells you a lot about how a system was built.

• Power conversion (the PCS / inverter). Batteries store DC; the grid runs on AC. Every charge converts AC to DC and every discharge converts it back, and each conversion step loses 1–3%. A modern power conversion system peaks above 98% efficiency but only in a narrow load band — run it lightly loaded and efficiency falls off. This is usually the single largest loss in an AC-coupled system.
• Ohmic / IR heating. Current flowing through cell internal resistance, busbars, contactors and cabling dissipates energy as heat, scaling with the square of the current. Double the current and you quadruple this loss — which is why C-rate matters so much (below).
• Auxiliary / parasitic loads. The cooling pumps, fans, cabinet HVAC, the BMS itself and the controller all draw power around the clock. On a hot day a liquid-cooling loop working hard can quietly eat 1–2% of throughput. These loads run whether or not the battery is cycling, so a system that sits idle at a high state of charge still bleeds.
• Self-discharge. Cells slowly lose charge sitting still. For LFP this is small — a fraction of a percent per month — but for a battery that charges in the morning and discharges at night it is a rounding error, while for one that charges seasonally it is not.

AC round-trip vs DC round-trip — where the marketing hides

This is the most important paragraph in the article. Round-trip efficiency depends entirely on where you put the meters, and there are two common boundaries:

• DC round-trip efficiency measures at the DC battery terminals — cells, BMS and internal wiring only. It excludes the inverter entirely. This number is flattering: 96–98% is normal because you have left the biggest loss out of the measurement.
• AC round-trip efficiency measures at the AC grid connection — the whole system, inverter and auxiliaries included. This is the number that matters, because it is the electricity you actually buy from the grid and the electricity you actually sell back. It is typically 3–5 points lower than the DC figure for the same hardware.

So a vendor can honestly print "96% round-trip efficiency" on a datasheet and be describing a system that returns 91% at the point where you get paid. Neither number is a lie; the DC one is just answering a question you did not ask. A datasheet that quotes an RTE without stating the measurement boundary — AC or DC, and at what C-rate and temperature — is hiding something. Always ask which terminals, and always compare AC-to-AC when you compare quotes.

How a few points compound into a decade of opex

Efficiency loss is a tax you pay on every kWh that passes through the battery, forever. Consider a 1 MWh commercial system cycling once a day at an electricity price of $0.15/kWh.

The gap between a 95% system and an 88% one is on the order of $37,000 in wasted energy over ten years on a single 1 MWh cabinet — before you count that the lower-efficiency system also dumps more heat, which loads the cooling system harder and can shorten cycle life. Scale that across a multi-megawatt site and RTE stops being a footnote. This is precisely why it feeds directly into LCOS (levelised cost of storage): every point of round-trip efficiency lowers the cost of every stored kWh across the asset's whole life. Two systems with identical cell prices and identical warranties can have materially different lifetime economics purely on the efficiency line.

Modern LFP numbers vs cheaper builds

Not all LFP systems land in the same place, and the spread comes almost entirely from system engineering rather than the cells.

The cells inside a $/kWh-optimised cabinet and a premium one may come from the same factory. The efficiency gap opens up in the inverter quality, the cooling control strategy, the busbar and connector design, and how tightly the cells were matched for internal resistance. A good energy management system also helps by scheduling charge and discharge into the inverter's high-efficiency load band instead of trickling power at its inefficient extremes.

How C-rate and temperature drag RTE down

The single RTE figure on a datasheet is measured under one specific set of conditions. Change the conditions and the real number moves — usually downward.

• C-rate. Because ohmic loss scales with the square of current, pushing a battery harder costs efficiency fast. A system that returns 94% at a gentle 0.5C (a two-hour discharge) may return only 90–91% at 1C and worse at 2C. A cabinet sized for a two-hour duty and then run flat-out on a one-hour duty will underperform its own datasheet — which is why the C-rate at which RTE was measured has to be stated. This is one reason high-cycling, high-power applications favour hardware built for the job rather than a lower-rated unit driven hard.
• Temperature. Cold raises internal resistance, so cold cells are less efficient and lose more energy to IR heating. Heat is the sneakier problem: high ambient temperatures make the thermal management system work harder, and that parasitic cooling load comes straight out of round-trip efficiency. A system that reads 94% on a mild bench day can slip a point or two on a 40 °C afternoon when the cooling loop is running flat out.

The takeaway: an RTE number without its conditions is close to meaningless. "94% at 0.5C, 25 °C, measured AC-to-AC" is an engineering claim. "Up to 96%" with no conditions is a marketing one.

How Hua Power measures round-trip efficiency

We quote round-trip efficiency the way we think it should be quoted — at the boundary where you actually pay for and sell electricity, with the conditions stated.

The measurement boundary

• Round-trip efficiency of 92–95% measured at the AC terminals on a 1-hour discharge — an AC-to-AC figure with the inverter and all auxiliary loads inside the measurement, not a flattering DC-only number
• Built on a standardized 3.2 V / 314 Ah large-format LFP cell, sorted at the production line into matched groups to keep the internal-resistance spread tight and IR losses low

The hardware that protects the number

• Liquid-cooled cabinets — the HC261P and HC522P lines — use demand-based cooling to cut the parasitic and thermal losses that punish high-cycling duty, so RTE holds up under a hard daily schedule instead of collapsing on hot days
• In-house BMS hardware and firmware with per-cell-group monitoring, keeping cells balanced and inside their efficient operating band
• Visual EMS scheduling that charges and discharges through the inverter's high-efficiency load range rather than its inefficient extremes

The range

• 17 standardized C&I SKUs from 64 kWh up to 1.2 MWh, so a system can be matched to the real duty cycle instead of being oversold on power and run inefficiently — or undersized and driven past its rated C-rate

If you are comparing two quotes whose efficiency numbers look close, ask each vendor the same three questions: AC or DC, at what C-rate, and at what temperature. The differences that decide a decade of operating cost live in those answers. The contact form below goes straight to our engineering team.

Round-trip efficiency is the tax you pay on every stored kilowatt-hour for the life of the asset. A couple of points sound trivial on a datasheet and add up to tens of thousands of dollars per megawatt over a decade — so always compare AC-to-AC, at a stated C-rate and temperature, and treat any RTE quoted without its measurement boundary as a number designed to be misread.