Backup power for a distribution centre: UPS vs generator vs BESS

The question isn't really "UPS or generator" — it's what each one is actually for, and most distribution centres end up needing both plus a view on whether battery storage belongs in the mix too. A UPS keeps the lights and the network up for the seconds and minutes it takes something else to take over. A generator keeps them up for the hours it takes for grid power to come back. Confuse the two roles and you either overspend on batteries trying to ride out a multi-hour outage, or underspend on a generator start-up buffer and watch your WMS servers reboot every time the grid blinks.

The decision gets harder once refrigeration, conveyor automation, and a warehouse management system with live inventory data are all sitting on the same site, because each of those loads has a different tolerance for an interruption. A dropped connection on an RF scanner is an annoyance. A cold chain excursion on temperature-controlled stock is a write-off. Sizing backup power starts with being honest about which loads actually belong in which category, not with picking a technology first.

Key Takeaways

• A UPS provides ride-through — instantaneous, uninterrupted power for seconds to tens of minutes — not sustained backup; its job is to bridge the gap until a generator starts or systems shut down cleanly.
• A diesel generator covers sustained outages for as long as fuel holds out, but takes roughly 10 to 30 seconds to start and reach stable output, which is exactly the gap a UPS is sized to cover.
• Battery energy storage systems (BESS) sit between the two: longer duration than a typical UPS, silent and emission-free, but ultimately capacity-limited unless recharged, unlike a generator that just needs more fuel.
• Sizing starts with separating critical loads (servers, network, refrigeration, safety systems) from non-critical ones (general lighting, non-essential equipment), because backing up everything at generator-grade cost is rarely the right answer.
• Generator capacity has to account for motor starting current on compressors, conveyor drives, and forklift chargers, not just steady-state running load, which is why generators are usually sized well above the facility's average draw.
• The UPS capacity calculator turns your connected critical load and desired ride-through time into a starting kVA figure before you get a generator quote.

What each technology actually covers

A UPS (uninterruptible power supply) sits between the grid and your critical loads, drawing from a battery bank the instant grid power drops, with no perceptible break in supply. Its defining constraint is duration: a UPS sized for a typical distribution centre server room or control panel might provide anywhere from a few minutes to 15–30 minutes of runtime at full load, depending on how many battery modules sit behind it. It is not designed to run a facility through a multi-hour outage — it's designed to keep sensitive electronics alive and stable for exactly as long as it takes for something else to pick up the load, or for an orderly shutdown to happen.

A generator is the sustained-outage answer, an engine-driven power source that can run for as long as fuel is available and resupplied, which makes it the only one of the three genuinely suited to outages measured in hours rather than minutes. Its limitation is at the start, not the finish: from the moment grid power fails, an automatic transfer switch (ATS) has to detect the outage, signal the generator to start, and the generator has to crank and stabilise before it can safely carry load. NFPA 110, the US standard most facility engineers still design against for standby power, sets a 10-second time-to-readiness for its strictest "Type 10" life-safety class: a mandatory 1-second delay to filter out momentary grid dips, followed by up to 9 seconds for the engine to start, stabilise, and transfer load (Cummins Inc., "10-Second Start: A Look at NFPA 110 and Time to Readiness Following Power Loss", retrieved 2026-09-08). Non-life-safety standby systems are commonly specified looser, in the 10-to-30-second range used elsewhere in this guide. That gap is not the generator failing; it's just how the mechanism works, and it's precisely the interval a UPS is sized to bridge so that servers and control systems never see the interruption at all.

Battery energy storage (BESS) is the newer option, and it occupies a middle ground. A BESS can be sized for far longer duration than a conventional UPS — multiple hours rather than minutes — while remaining silent, emission-free, and free of fuel logistics, which matters where diesel storage is restricted or where a facility sits close enough to other buildings that generator noise and exhaust are a real constraint. Its limitation is the mirror image of a generator's strength: capacity is fixed by the size of the battery bank, and once it's depleted, it needs to recharge from the grid or from solar before it can do it again. A generator with a fuel truck can run indefinitely; a BESS on its own cannot.

Building the load profile first

Before sizing anything, list what actually needs to stay powered and for how long. In a typical UAE distribution centre that usually breaks down into a handful of categories: refrigeration and cold storage compressors, where even a short excursion outside temperature range can spoil stock; the WMS, RF network, and any servers or edge devices that inventory accuracy depends on; life-safety systems — fire detection, emergency lighting, access control at exit routes — which often carry their own regulatory backup requirements independent of your operational choices; and then the larger, harder call: whether conveyor and sortation automation, dock levellers, and general work-area lighting need to keep running through an outage, or whether it's acceptable for picking to pause until power returns.

That split matters because backing up everything at generator-grade reliability is expensive relative to backing up only what genuinely can't tolerate an interruption. Many facilities land on a tiered approach: UPS coverage on IT, network, and critical control loads for clean, instant ride-through; generator coverage on refrigeration, safety systems, and enough of the operational floor to keep receiving and shipping moving during a multi-hour event; and everything else accepted as a planned pause.

Sizing logic that holds up

UPS sizing starts from the connected critical load in kVA, adjusted for the actual power factor of that equipment, and then multiplied out against how many minutes of autonomy you want — enough to cover your generator's realistic start time with margin, or enough to cover a full graceful shutdown if there's no generator behind it at all. Extending runtime is mostly a matter of adding battery cabinets, so the real design decision is choosing the target duration honestly rather than defaulting to whatever the vendor's standard configuration happens to be. The UPS capacity calculator takes your connected load and target runtime and returns a starting kVA and battery configuration to quote against.

Generator sizing has a different trap: steady-state running load understates what the generator needs to handle at the moment of starting. Refrigeration compressors, conveyor motors, and forklift battery chargers all draw a surge of starting current well above their running current, and if several of them try to restart simultaneously when the generator picks up the load, an undersized unit can stall or trip rather than carry it. Generators are typically specified with headroom above calculated running load specifically to absorb that starting surge, and a facility with several large motors on a shared circuit needs that margin checked against actual equipment nameplate data, not just a rule-of-thumb percentage. Fuel tank sizing then follows from the target runtime you're planning for — commonly enough for a full day's outage as a baseline, more where redundancy (N+1) matters and grid restoration times in your area have historically run long.

Worked example: a mid-sized UAE distribution centre with 40kVA of critical IT/network load, 120kVA of refrigeration compressors, and 60kVA of conveyor/dock equipment might land on: a 50kVA UPS sized for 15 minutes of autonomy on the IT/network load alone (enough to bridge a 30-second generator start with wide margin, or complete a clean shutdown if the generator fails to start at all); a 250kVA generator covering refrigeration, safety systems, and conveyor/dock load, sized above the 220kVA combined running total to absorb compressor and conveyor-motor starting surge; and a fuel tank sized for a full day's runtime at that load as the baseline. The UPS and generator numbers here are illustrative. Run your own connected-load figures through the UPS capacity calculator rather than reusing this example's kVA values.

Making the actual decision

There's no single right answer across UPS, generator, and BESS — the right combination depends on how tolerant your specific loads are of a gap in power, how often and how long outages actually run in your area, and how much noise, emissions, and fuel logistics you're willing to carry on site. A cold-chain facility with automation and a live WMS generally ends up with all three roles filled by something: UPS for the instant bridge, generator for sustained duration, and increasingly BESS layered in either to extend UPS-grade quiet backup or to reduce how often the generator needs to run at all. A dry-goods facility with simpler equipment and more tolerance for a pause may reasonably conclude that UPS on the network gear plus a modest generator is enough, without adding battery storage into the mix at all. If you'd rather have that UPS/generator/BESS mix scoped against your actual load profile by a specialist than work it out entirely in-house, WiserMonks' critical facility power backup service does exactly that sizing exercise for a distribution centre's specific critical-load mix.

Frequently asked questions

Can a UPS replace a generator entirely?

Not for a facility-wide outage. A UPS is built for ride-through measured in minutes, sized by battery capacity that runs down and needs recharging; a generator is built for sustained duration measured in hours, limited mainly by fuel supply. The two solve different problems and most distribution centres with meaningful critical load need both.

Why does the generator take time to start, and does that matter?

Yes, it matters, and it's normal. Between the moment grid power fails and the moment a generator is stable enough to carry load, an automatic transfer switch has to detect the failure and the generator has to crank and stabilise — typically on the order of 10 to 30 seconds. Anything that can't tolerate even that short a gap, such as servers or control systems, needs a UPS in front of it regardless of how good the generator is.

Does a BESS make sense without a generator?

It can, particularly where outages are typically short and frequent rather than long and rare, or where noise and emissions restrictions rule out a diesel generator on site. The trade-off is that a BESS on its own is still capacity-limited — it will not carry a facility through an extended multi-hour or multi-day outage the way a fuelled generator will.

How much backup runtime should we actually plan for?

Start from your own outage history and your grid provider's track record in your specific area, not a generic industry figure. Facilities with critical cold chain or automation typically plan generator runtime in terms of a full day as a baseline, with UPS autonomy sized only to bridge the generator's start-up interval or to allow a clean shutdown where no generator exists.

The bottom line

UPS, generator, and BESS aren't competing options for the same job — they cover different lengths and qualities of outage, and the right backup power setup for a distribution centre is usually a combination sized against an honest list of which loads can't tolerate an interruption at all, which can tolerate a brief one, and which can simply wait.

This guide was reviewed and verified on September 6, 2026.