Battery, UPS or Generator? Backup Power for a Business Site

When the grid drops, three different pieces of equipment can keep a business running: an uninterruptible power supply (UPS), a standby generator and a commercial battery designed for backup. They are often discussed as if they were rivals. They are not. Each one answers a different question: how fast the power must come back, how long it must last, and what the equipment does for the other 364 days of the year when the grid is fine.

This guide sets out what each option really does, the design details that decide whether a battery can provide backup at all, and a simple way to size the backup element before you talk to an installer. For the wider decision, see our guide to buying commercial storage.

First, a trap: most grid-connected batteries switch off in a power cut

A behind-the-meter battery bought to cut peak charges is, by default, a grid-following system. When the network fails, its protection is required to disconnect it from the grid. That is loss-of-mains protection, part of the G99 connection requirements, and it exists so the battery cannot energise a network that engineers believe is dead. Unless the system was specified for backup, it simply stops, and your site goes dark along with everyone else’s.

A battery only provides backup if it was designed to. That takes three things:

• An inverter or power conversion system that can form its own grid (often called grid-forming or islanding-capable), so it can run your loads with the network disconnected.
• An automatic transfer switch or equivalent isolation, which separates the site from the grid before the battery takes over and reconnects it safely afterwards.
• An essential-loads board, so the battery carries the loads that matter, such as IT, refrigeration, security, emergency lighting and a pick line, rather than the whole building.

If a quote for “battery storage with backup” does not name all three, ask. And the backup arrangement belongs in the G99 application, not bolted on afterwards.

What each option does

Uninterruptible power supply (UPS)

A UPS sits between the mains and a specific set of loads, usually IT, comms and control systems. An online (double-conversion) UPS supplies those loads through its own electronics all the time, so a mains failure causes no break at all. That is why server rooms and payment systems use them.

The trade-off is runtime and scope. A UPS is normally sized to carry a limited load for minutes, long enough to ride through a short interruption or shut systems down cleanly. It does nothing on a normal trading day except wait.

Standby generator

A diesel or gas standby generator gives the longest runtime: as long as there is fuel, it keeps running. That matters for sites where an outage could last many hours, such as hospitals, cold stores with high-value stock and sites with life-safety systems.

Generators are not instant. An automatic mains failure panel detects the outage, starts the engine and transfers the load once the set is up to speed, which takes seconds, not milliseconds. So sites with sensitive IT pair a generator with a UPS. Generators also need regular test runs under load, fuel management and servicing to stay reliable. In England, a diesel tank holding more than 200 litres outside a building falls under the Control of Pollution (Oil Storage) (England) Regulations 2001, which set containment requirements. Between faults, a generator earns nothing.

Backup-capable commercial battery

A battery designed for backup can pick up essential loads very quickly. How quickly depends on the design: some architectures switch in a short break as the transfer switch operates, others keep critical loads on the battery’s electronics continuously. The runtime is fixed by the usable energy (kWh) you reserve for outages.

The difference is what it does the rest of the year. The same battery works every day on peak-shaving and load shifting, cutting red-band distribution charges and moving cheap overnight energy into the expensive afternoon. It is the only one of the three options that pays for itself when nothing goes wrong.

The trade-off nobody mentions: reserved charge

A battery cannot spend energy on peak-shaving and hold the same energy back for a power cut. Every kWh you reserve for backup is a kWh that cannot earn money that day. A 30% reserve means only 70% of the usable capacity works on savings.

That is not a reason to avoid backup. It is a reason to size the reserve deliberately: protect the loads that genuinely cost money when they stop, for the time you genuinely need, and let the rest of the capacity earn. A good feasibility model shows the saving both with and without the reserve, so you can see what resilience costs.

How to size the backup element

The arithmetic is simple enough to do before you speak to anyone:

1. List the essential loads and their power in kW. Use nameplate ratings or, better, logged demand for the circuits concerned.
2. Decide the runtime you need. Minutes to shut down cleanly? One hour to finish a shift? Four hours to protect chilled stock?
3. Multiply. Essential load (kW) × runtime (hours) = the usable energy (kWh) to reserve. For example, 40 kW of essential load for 2 hours means reserving 80 kWh of usable capacity.
4. Allow for losses and limits. Inverter efficiency and the cells’ permitted depth of discharge mean the installed capacity must be larger than the usable reserve. Your installer should show this step.
5. Check the power rating as well as the energy. The inverter must supply the essential loads’ combined peak, including motor start-up currents, not just their average.

If step 3 produces a reserve that would swallow most of the battery, the honest answer is usually a battery for short outages and daily savings, with a generator behind it for long outages.

When a site needs more than one

• Offices and light commercial sites with IT: a UPS for the server room is often enough. A battery adds daily savings and covers wider essential loads for a short outage.
• Distribution and fulfilment sites: a backup-capable battery covering the warehouse management system, pick lines and dock doors handles the common short interruption, with a generator only if long outages would stop despatch for a critical customer.
• Cold stores and food production: runtime drives the answer. A battery handles short interruptions and the daily peak; stock value usually justifies a generator for long ones.
• Life-safety systems: follow the specific standards for those systems. Battery storage does not replace dedicated emergency lighting or fire-system power supplies.

Questions to ask before you sign

• Is the system grid-forming or islanding-capable, and what is the transfer arrangement and expected break time?
• Which circuits are on the essential-loads board, and what reserve (kWh) is held for them?
• How much daily saving does the reserve cost in the feasibility model?
• Is the backup arrangement declared in the G99 application, with loss-of-mains protection?
• Is the design to the IET Code of Practice for Electrical Energy Storage Systems, and has your insurer seen it?

Location matters too, because your distribution network operator’s red-band times shape the daily savings the battery earns between outages. Our guide to backup power for Milton Keynes businesses works through this for one of the UK’s densest distribution markets, and commercial battery backup power covers the design in more depth. For your own site, the feasibility summary below models the saving with and without a backup reserve.

This guide is independent information published by SEO Dons Ltd. We are not an installer; we model the numbers and, if storage stacks up, introduce you to MCS-certified installers.