The popular claim: “APC Smart-UPS and Schneider Galaxy VS are essentially the same company, so runtime and efficiency are equivalent — just pick the one that fits your rack.” That sounds reasonable until you map the actual load vs runtime curve for a 10 kW data hall fed by a 10 kVA UPS. The myth conflates topology with power delivery and ignores how the single most influential variable — real power factor (PF) of the load — hollows out runtime predictions. This article debunks the equivalence assumption by following one variable (load PF) through the specification chain.
Myth #1: “10 kVA = 10 kW, so runtime curves transfer directly”
Myth
A 10 kVA UPS delivers 10 kW; the runtime table for full load applies equally whether the load is a server farm or a lighting panel.
Reality
Output power factor (PF) determines how many real watts the UPS can supply. APC Smart-UPS Online SRT at 10 kVA has a unity output PF (1.0), so it can deliver 10 kW [apc-ups]. Schneider Galaxy VS at 10 kVA is specified for 3-phase loads and its output PF is typically 0.9 lagging (10 kVA → 9 kW), unless derated [schneider-ups]. A 10 kW real load on the Galaxy VS would require an 11.1 kVA unit; running it at 10 kW on a 10 kVA unit forces the inverter into current limit, triggering a bypass or overload — runtime becomes zero.
Mechanism: The UPS inverter is rated in VA (volts × amps), but the battery energy is stored in watt-hours. A load with a PF of 0.9 draws more current per real watt than a load at PF 1.0. Higher current → higher I²R losses in the inverter and battery interconnects → less usable energy for the same battery Wh. The consequence is a runtime that can be 10–15% shorter than the VA-based table suggests when the load PF is below the UPS’s rated PF.
Worked consequence: Suppose a 10 kW IT load (PF 0.95) is connected to a Schneider Galaxy VS rated 10 kVA / 9 kW. The UPS sees 10 kW / 0.95 = 10.53 kVA — beyond its 10 kVA rating. The unit either alarms or transfers to bypass. Runtime = zero. If you instead sized a 12 kVA Galaxy VS (12 kVA × 0.9 = 10.8 kW), runtime at 10 kW (95% load) is about 4–6 minutes on standard battery [schneider-ups]. Under the same load with an APC SRT 10 kVA (10 kW capacity at PF 1.0), the load is at 100% capacity, runtime ~5 minutes [apc-ups]. The difference is marginal at full load, but the real trap is the sizing requirement: the Schneider unit must be upsized by one frame to handle the same real load, adding cost and footprint.
Reversal: If the load is purely resistive (PF 1.0, e.g., lighting or resistive heaters), the 10 kVA Galaxy VS will deliver 9 kW — the load is 9 kW, not 10 kW — so runtime is slightly longer than the APC at 10 kW. But for any data-center load with typical server PFC (PF 0.95–0.99), the APC’s unity PF gives a tighter fit.
Myth #2: “Double-conversion efficiency is the same across the range”
Myth
Online UPS efficiency is about 95–96% regardless of load level; runtime is battery capacity divided by load watts, adjusted for efficiency.
Reality
Efficiency varies significantly with load percent. The Galaxy VS in double-conversion mode is up to 97% at every load level, but the APC Smart-UPS Online SRT is typically ~94% at 50% load and ~95% at 100% load, with Green Mode up to 98% only when bypassing the inverter [apc-ups][schneider-ups]. A 2–3% efficiency delta at half load means the same battery yields ~2–3% more runtime on the Galaxy VS.
Mechanism: Double-conversion efficiency is dominated by inverter and rectifier losses. The Galaxy VS uses a 4-level inverter topology that maintains high efficiency across a wide load range [schneider-ups]. APC’s SRT uses a conventional 2-level IGBT design; its efficiency peaks near full load and drops faster at light loads (
Worked consequence: Assume a 10 kVA Galaxy VS at 50% load (5 kVA, 4.5 kW) with 97% efficiency means the input power to the rectifier is 4.64 kW. The battery energy required to support 4.5 kW for 10 minutes is ~0.75 kWh (ignoring battery efficiency). At 94% efficiency (APC SRT), the input power is 4.79 kW → same 10-minute runtime needs ~0.80 kWh. The 5.5% extra battery energy needed (0.05 kWh) translates to roughly 30 seconds less runtime from the same battery capacity. Over a 30-minute runtime expectation, you lose ~1.5 minutes. This is material when sizing for orderly shutdown.
Reversal: If the load is consistently at 80–100% of UPS rating, both units converge to similar efficiency (~95–96%), and the runtime difference shrinks to
Myth #3: “Battery runtime is simply battery Ah × voltage ÷ load watts”
Myth
Runtime follows a linear curve: double the load, half the runtime. The datasheet table is reliable for any real load.
Reality
Battery capacity derates non-linearly with discharge current (Peukert’s law). A UPS with a larger battery bank can sustain higher discharge currents with less capacity loss. The Galaxy VS can be configured with up to 5 battery strings in parallel [schneider-ups], reducing the per-string discharge rate and improving effective capacity. A single-string APC SRT at the same kVA rating will have a steeper runtime decline at high loads.
Mechanism: Lead-acid batteries (common in both) lose capacity as discharge current increases. A battery rated 100 Ah at the 20-hour rate (C/20) may deliver only 70 Ah at the 1-hour rate (C/1). If the UPS is sized with one string, a 10 kW load forces a high C-rate, reducing usable capacity by 20–30% compared to the linear prediction. With multiple strings, each battery sees a lower current, so Peukert losses shrink.
Worked consequence: Consider a 10 kVA Galaxy VS with two battery strings (each 144 V, 9 Ah at C/20 → ~18 Ah total). At 9 kW load (C-rate ≈ 9 kW / (144 V × 2) ≈ 31 A per string → ~2.2C), the effective capacity might be ~60% of rated → ~10.8 Ah. Runtime ≈ (10.8 Ah × 288 V) / 9000 W ≈ 0.35 h = 21 minutes. A single-string APC SRT at 10 kVA (same 144 V, 9 Ah) sees 9000 W / 144 V = 62.5 A → 6.9C → effective capacity ~40% → 3.6 Ah → runtime ≈ (3.6 × 144) / 9000 = 0.058 h = 3.5 minutes. The multi-string advantage is dramatic at high loads. At half load (4.5 kW), the C-rate is lower and the gap narrows: Galaxy ~12 minutes, APC ~10 minutes.
Reversal: If the load is consistently below 30% of UPS rating (e.g., oversized UPS), Peukert losses are small and runtime differences between single and multiple strings diminish. Also, lithium-ion batteries (optional on both) have negligible Peukert effect, making the string-count advantage irrelevant if Li-ion is used.
Myth #4: “Runtime is all about the battery — the UPS electronics don’t matter”
Myth
Once you have a battery, the UPS just inverts DC to AC; any extra losses are too small to matter.
Reality
Internal auxiliary loads (fans, control logic, display) and inverter idle losses can consume 200–400 W even with zero output load. The Galaxy VS in eConversion mode (bypass with inverter on standby) draws ~100 W auxiliary, while a double-conversion SRT at idle might draw 250–350 W [apc-ups][schneider-ups]. For a small load (e.g., 1 kW), the UPS self-consumption can eat 25–35% of the battery, destroying runtime.
Mechanism: Online UPS must keep the inverter running continuously. The inverter’s switching losses and the cooling fans are a fixed overhead. At low loads, this overhead is a larger fraction of total load. The Galaxy VS’s eConversion mode (approved for data centers) keeps the inverter in standby and powers the load via the bypass, but can transfer to double-conversion in
Worked consequence: A 1 kW load on a 10 kVA UPS: Galaxy VS in eConversion → total DC load = 1 kW + 0.1 kW = 1.1 kW. APC SRT in double-conversion (Green Mode off) → total load = 1 kW / 0.94 + 0.3 kW ≈ 1.36 kW. The APC draws 24% more from the battery → runtime is 24% shorter. With a 30-minute battery, you get ~23 minutes on APC vs ~27 minutes on Galaxy VS — a meaningful gap for orderly shutdown.
Reversal: At loads above 50% of UPS rating (e.g., 5 kW on a 10 kVA unit), the auxiliary overhead is
If your real load is above 80% of the UPS’s rated output power (in kW) and the load PF is close to unity (≥0.95), the Schneider vs APC runtime difference is under 10% — choose on other criteria (service, software). If your load is below 50% of rating or the load PF is below 0.9, the Galaxy VS will deliver 15–25% more runtime from the same battery bank due to higher efficiency and multi-string flexibility. If you must size for a fixed runtime (e.g., 10 minutes at full load), the Galaxy VS may allow one fewer battery string, reducing cost and footprint.
Topology/standards per the cited standards; all product ratings are manufacturer-stated values from the cited datasheets, current to 2026-06; derived/illustrative figures are labelled as such. This is not an independent head-to-head test. Schneider Electric is a brand affiliated with this site; competitor names are used for identification only.
- IEC 62040-3 UPS classification (transfer time, topology)
- APC Smart-UPS Online SRT datasheet (topology, PF, efficiency, Green Mode)
- Schneider Galaxy VS datasheet (topology, efficiency, eConversion, battery config)
- Schneider eConversion white paper (auxiliary power, transfer modes)