Sizing storage for night-time load, not headline capacity

Key Takeaways

• A battery's headline kWh rating is its total energy capacity, not the usable energy available for your night-time load, since usable capacity is reduced by both depth-of-discharge limits and inverter efficiency losses.
• Required storage capacity should be calculated as: overnight load (kW) × required hours ÷ usable depth of discharge ÷ inverter efficiency, not simply matched to a round battery size that "sounds about right."
• Inverter losses alone typically consume 10-15% of rated capacity in real-world use, and shallower depth-of-discharge targets (to protect cycle life) further reduce the energy genuinely available each night.
• Night-time load isn't flat: sizing to the average overnight draw rather than its actual peak-and-trough shape risks a shortfall exactly when a spike (a compressor cycling on, an appliance starting) occurs late in the discharge window.

A battery advertised as "10 kWh" doesn't deliver 10 kWh to your night-time load. Between depth-of-discharge limits protecting cycle life and inverter conversion losses, the energy actually available is meaningfully less than the number on the spec sheet, and sizing a system against the headline figure rather than the real usable capacity is one of the most common ways a solar-plus-storage system underperforms its own design intent.

The gap between rated capacity and usable capacity

Two separate factors reduce a battery's rated capacity down to what's genuinely available for load. First, running a battery to a shallower depth of discharge protects cycle life (see the companion article on depth of discharge and cycle life for the mechanics), which means a portion of the rated capacity is deliberately held in reserve rather than used. Second, converting stored DC energy to usable AC power through an inverter isn't lossless: high-quality sine-wave inverters typically run at 90-95% efficiency, meaning 5-10% of the energy is lost in conversion alone, and real-world system losses (wiring, thermal derating, standby draw) commonly bring total losses to 10-15% of rated capacity (Oukitel Power, portable power station calculator guide, retrieved 2026-09-10). Stack both factors and a battery rated at 10 kWh with an 80% DoD target and 90% inverter efficiency delivers roughly 7.2 kWh of genuinely usable energy to load, not 10.

The correct sizing formula, worked backwards from load

The right approach is to size backwards from the actual overnight load, not forwards from a battery's headline rating: required Wh (or kWh) equals load watts multiplied by required hours of coverage, divided by usable depth of discharge, divided by inverter efficiency (Outbound Power, selecting the right size portable power station, retrieved 2026-09-10). A facility with a genuine 2 kW overnight draw needing 8 hours of coverage, at 80% usable DoD and 90% inverter efficiency, needs roughly 2 × 8 ÷ 0.8 ÷ 0.9 ≈ 22.2 kWh of rated battery capacity, not the 16 kWh a naive load × hours calculation would suggest. Run your own overnight load profile through the power bank/battery runtime calculator using this formula rather than rounding to a battery size that looks approximately right.

Why average load isn't the number to size against

Night-time load is rarely flat: a compressor cycling on, a pump starting, or an appliance's startup surge can spike demand well above the average overnight draw, sometimes 2-3x the running wattage for motor-driven loads specifically (Renewable Outdoors, portable power station watt-hour calculator, retrieved 2026-09-10). Sizing a battery to the average overnight load, rather than accounting for its actual peak-and-trough shape, risks the system running out of usable capacity or hitting an inverter power limit exactly during a late-night spike, which is precisely when it's least convenient to discover the shortfall. Model the load curve's shape, not just its average, particularly if any equipment on the circuit has a motor or compressor.

What oversizing costs, and why it's the safer failure mode

Oversizing a battery relative to actual overnight need costs more upfront capital and, ironically, can also mean running the battery at a shallower DoD than necessary on most nights, which is not itself a problem (it extends cycle life) but does represent capital tied up in capacity rarely used. Undersizing, by contrast, means the system runs dark, drawing from the grid or a generator earlier than planned, which defeats the purpose of the storage investment on the nights it matters most. Between the two failure modes, sizing conservatively (erring toward more capacity) is the safer default when the load profile carries genuine uncertainty, but that conservatism should be a deliberate choice, not a substitute for actually modelling the real overnight load curve. Once the required usable capacity is calculated, checking it against the battery energy storage systems range shows which off-the-shelf unit sizes actually match that figure, rather than rounding to whatever's readily available.

Frequently asked questions

Why does a "10 kWh" battery not deliver 10 kWh of usable energy?

Depth-of-discharge limits (to protect cycle life) hold back some rated capacity from routine use, and inverter conversion losses (typically 5-10%, more with total system losses) reduce the energy that actually reaches the load. Combined, these commonly reduce usable energy to 70-85% of the rated figure.

Should I size storage to my average overnight load or its peak?

Neither alone. Size total capacity to cover the full overnight energy need (which relates to average load over the full duration), but confirm the battery's power rating (kW) separately against the load's actual peak, including motor-driven equipment startup surges, since a system can have enough total energy but still trip on an instantaneous power spike.

Is it better to oversize or undersize a storage system if I'm unsure of the exact load?

Oversizing is the safer failure mode: it costs more capital and may mean running at a shallower DoD than strictly necessary, but it doesn't leave the facility without power. Undersizing risks a shortfall precisely during the load spikes that occur late in the overnight window, which is the worst time for it to happen.

The bottom line

A battery's rated kWh figure and the energy actually available to power your night-time load are two different numbers, separated by depth-of-discharge margin and inverter losses that commonly total 15-30% of rated capacity. Size backwards from your actual load curve, including its peaks, through the full formula, not forwards from a battery size that looks like it should be big enough.

Figures were verified on 10 September 2026 against published battery and power station sizing methodology sources. Actual usable capacity depends on specific battery chemistry, inverter model, and system design; confirm manufacturer-specific efficiency and DoD figures before finalising a sizing calculation.