What exactly is UPS? ( Part 2)

III. The Three UPS Brothers: What are the differences?

Based on their operating methods, UPS systems are divided into three types. To use a restaurant analogy:

Backup type – the spare tire type. Normally, the equipment runs directly on mains power, with the UPS on standby. It switches over only the instant of a power outage, a process that takes a few milliseconds. Like a spare tire in a car: you don’t use it normally, you only change it when it blows out, and the car pauses briefly during the change. This is inexpensive and sufficient for home and small office use.

Online interactive system – an upgraded backup. It monitors the mains power supply, adjusts it slightly if there’s any voltage fluctuation, and then takes over when the power is cut off. The “pause” time is even shorter. A middle-ground option.

Online (Double Conversion) – Central Kitchen. When mains power comes in, regardless of its quality, it’s first rectified into DC (for reprocessing), then inverted into AC (for output). Every bit of electricity the equipment consumes is reprocessed by the UPS, always clean and seamless. Power outage? The battery seamlessly replenishes the power, without even a pause.

Data centers only use the third type. There are two reasons: the switching time is close to zero; and the electricity is always clean.

IV. Understanding the parameters: Three words are enough

Anyone who has bought a power strip knows the term “rated power.” The parameters for a UPS are logically the same, consisting of only three:

Capacity (kVA / kW): How many devices it can power. A UPS in a computer room can range from tens of kVA to MW—it’s not just a “power bank,” it’s an entire wall.

Battery life: How many minutes it can last without mains power. Note: Longer isn’t always better. Server rooms are typically designed for 5-15 minutes. Why so short? Read on.

Switching time: How fast it is from a power outage to being back up. For online systems, it’s 0, which is why data centers choose it.

V. Three most common misconceptions, easily dispelled.

Myth 1: “The bigger the UPS, the better.” Wrong. The UPS capacity must match the load. Buying a larger one is just taking up space and costing more money, and the battery configuration will also be inflated. Choosing the right UPS is a matter of cost-effectiveness.

Misconception 2: “The longer the battery life, the better.” This is the most common mistake. UPS batteries in data centers are only designed for a few minutes, not because they’re stingy, but because they don’t need to be longer—because there’s a diesel generator behind the UPS. The correct procedure after a power outage is: UPS takes over immediately (0 seconds) → Diesel generator starts within ten seconds → Diesel generator stabilizes and takes over the entire system → UPS gracefully exits to recharge its batteries. Therefore, the UPS is the first runner in a relay race; its task isn’t to run the entire race, but to smoothly hand over the baton. Battery life is the margin left for this handover.

Myth 3: “UPS systems rarely break down, so you don’t need to worry about them.” While the UPS unit itself is indeed robust, the battery is a consumable. Lead-acid batteries need replacing every few years, and they die even faster at high temperatures. During my years maintaining server rooms, the majority of UPS-related alarms were never from the unit itself, but from the battery: capacity degradation, individual cell malfunctions, and loose connections. Anyone who underestimates the battery is setting themselves up for failure.