A containerized commercial system is usually introduced with two datasheet figures: rated energy and cycle life. Neither explains how it works. The number that does is the round-trip efficiency measured at the point of connection, and for a system of this class it settles near 88 percent. The whole architecture exists to hold that figure up, and every stage in the chain takes a slice of it.
Where the 88 percent is lost
A commercial battery does not store and return energy directly. It converts it several times, and each conversion has a cost. Taking 1,000 kWh drawn at the meter through one full charge and discharge cycle, with the stage efficiencies assumed for this illustration:
Worked example. Stage efficiencies are assumed values chosen to make the chain traceable, not measurements of a specific unit. Published product specifications state rated energy and cycle life; round-trip efficiency at the point of connection is a separate figure and has to be requested.
Read the last row again, because it is the one that surprises people. A system that is rated at 1,000 kWh returns 881 kWh at the same meter. The battery is the largest single loss, but the two conversion stages around it and the equipment that keeps the cells at temperature together take more.
Rated energy is the figure a datasheet is allowed to quote, and the 40GP container in Ruibit's published specification carries 2.089 MWh and a cycle life of at least 8,000, neither of which is the number that reaches the load.
What the number means when a quotation quotes capacity
Nameplate capacity is not delivered capacity. A 1,000 kWh system at a 90 percent depth-of-discharge convention and 88 percent round-trip efficiency delivers around 792 kWh to the load per full cycle. Three multiplications turn a nameplate into an energy bill, and buyers who skip the third one are the ones surprised in year two.
Two published project records show the scale this arithmetic governs: a containerized 1 MW / 2 MWh system at an industrial park in the Netherlands , completed in May 2026 with two 500 kW conversion units, and a 5 MWh solar plus storage installation in Zambia , completed in February 2026.
If the quotation states a single round-trip figure with no load point and no measurement boundary, it is a maximum, and the energy that actually arrives will be lower. Which means quotations are only comparable once they say three things: the load point the efficiency was measured at, the boundary it was measured between, and the depth-of-discharge convention behind the capacity figure.
What the number does not mean
Round-trip efficiency is not a property of the battery. It is a property of the whole chain, measured between two points that the buyer chooses. Measured at the PCS terminals it looks better than measured at the site meter, because the transformer, the cabling and the auxiliary load all sit outside the first boundary. Same equipment, two different numbers, neither of them dishonest.
It also does not stay constant. The 97.5 percent in the table is an operating-point value, not the peak figure a datasheet is entitled to print. A power conversion system is most efficient somewhere in the middle of its range and noticeably worse at the light loads that make up much of a commercial duty cycle.
When the number moves
Three things move it, and all three are fixed early enough that they are worth deciding deliberately.
Part load moves it most. If the duty cycle is dominated by operation below half power, the efficiency figure that matters is the one at 30 percent load, and the rated-load figure tells you very little.
Ambient temperature moves it through the auxiliary load. Above roughly 35 degrees C the cooling equipment works harder, and if the auxiliary load is measured inside the supplier's boundary, a sustained housekeeping load can cost more than a one-point efficiency gain.
Depth of discharge moves it because deeper cycles spend longer in the region where internal resistance is highest. And the cooling architecture, air or liquid, is the one of the four that cannot be revisited later, because it is welded into the enclosure before the system ships.
The second number you need
Round-trip efficiency answers how much energy survives the journey. It says nothing about how much energy the system consumes while doing nothing, which is the next number worth asking for.
Ask what the auxiliary consumption is at rated load, at part load, and in standby, in kilowatts rather than percentages. An answer of negligible, or included in the efficiency figure, means the number has never been measured at the boundary you are buying at. Ask for round-trip efficiency at 30 percent load and at 100 percent load, and ask where the boundary sits in both cases. Those three questions, answered in writing, turn an efficiency claim into a delivered-energy figure you can put into a financial model.
For the equipment that produces these losses, the component-level breakdown and the system-level picture of what a C&I BESS is are the two places to start, and the published specification for the 20GP container unit is where the nameplate figures used here come from.
FAQs
1. How does a commercial battery energy storage system work?
It converts energy rather than storing it directly. Power passes through a transformer and cabling, a power conversion system, the battery, the conversion system again, and the auxiliary equipment that keeps the cells at temperature. Every stage takes a share.
2. What is round-trip efficiency in a BESS?
The share of energy drawn at the meter that returns to the meter after one full charge and discharge cycle. In the worked example here, 1,000 kWh in returns 881 kWh, about 88 percent.
3. Why is delivered energy lower than nameplate capacity?
Nameplate capacity is multiplied by two more factors. A 1,000 kWh system at 90 percent depth of discharge and 88 percent round-trip efficiency delivers roughly 792 kWh per cycle.
4. Does round-trip efficiency change over time?
It is not constant even on day one. It falls at part load, drops as ambient temperature rises and the cooling equipment works harder, and degrades with the cells.
5. Should I compare quotes on round-trip efficiency?
Only if both state the load point and the measurement boundary, because a figure measured at the PCS terminals looks better than one measured at the site meter. Ask also for auxiliary consumption in kilowatts at rated load, part load and standby. An answer of negligible means it was never measured at the boundary being bought.
