UPS Planning for Critical Factory and Office Loads: Size the Gap, Not the Outage
Many factory and office UPS problems start with a purchase made by kVA and habit: one large unit, an hour of battery "to be safe", and every socket in the building connected to it. Then a voltage dip still trips a line, or the batteries are found flat on the day the DG fails to start. A UPS is a bridge: it carries the loads that cannot ride through a disturbance until the DG or the grid takes over, or until those loads have shut down safely. Planning one starts with a schedule of those loads and the time each must be carried.
The UPS runtime calculator screens energy and runtime for a given load; this guide covers what to put into it and what it cannot decide.
Build the load and autonomy schedule
List every load that might go on the UPS, measured rather than taken from nameplates, and decide for each how long it must be carried and why.
Load and autonomy schedule checklist
| Record for each load | What to capture | Why it matters |
|---|---|---|
| Load and location | PLC panel, server rack, network switch, SCADA computer, instrument, access control | Decides between one central UPS and distributed units |
| Measured kW and kVA | Power-analyser readings at typical and peak operation | Both UPS limits apply |
| Inrush and crest factor | Transformers, motors and power supplies at switch-on | High inrush pushes the UPS onto bypass |
| Ride-through needed | None, milliseconds, seconds or minutes | Separates no-break loads from loads that can restart on DG |
| Autonomy needed, and why | DG start and transfer time with margin; controlled shutdown time; time to save data | The battery follows from this, not from a round number |
| Supply behind the UPS | Whether a DG exists, and its measured start-and-transfer time | Without a DG, autonomy equals safe shutdown time |
| Redundancy need | Can the process tolerate a UPS fault? | Parallel, modular or N+1 arrangements |
| Growth | Additions expected on the same UPS | Headroom without overbuying |
- No-break loads — PLCs and control panels, servers and network equipment, critical instrumentation, controllers that lose position or data. These belong on the UPS.
- Short-break loads — motors, compressors, HVAC and general lighting restart when the DG picks up. Putting them on the UPS multiplies its size and its battery for little benefit.
- The awkward middle — drives on processes that must not stop. The answer may be a UPS on the drive's control supply only, a ride-through option offered by the drive manufacturer, or a UPS on the whole machine; decide machine by machine with the OEM.
- Life-safety systems — fire alarm panels, emergency lighting and similar systems have their own standby-power requirements under fire-safety rules. Do not treat a general-purpose UPS as their compliance route without confirming it with your fire consultant.
kW and kVA: two limits, both binding
A UPS has a kVA rating and a kW rating (kVA × its rated output power factor). The load must sit within both:
- Measure kW and kVA together. Adding up nameplates overstates some loads and misses the power factor of others.
- Modern IT power supplies can present a leading power factor at light load, and some UPS models derate for it. Check the manufacturer's output capability chart.
- Non-linear loads with a high crest factor, and inrush from transformers or motors, can push the UPS onto its static bypass. Keep heavy-inrush loads off the UPS or size for them.
- Leave headroom for growth and for redundancy, but avoid running a large unit lightly loaded for years: efficiency at low load and the capital tied up in batteries both suffer.
Topology and bypass
IEC 62040-3:2021 classifies UPS output performance. Three families matter in practice:
| Classification | How it behaves | Typical use |
|---|---|---|
| VFI — online double conversion | Load fed continuously through the rectifier and inverter; output independent of input voltage and frequency | PLCs, servers and instrumentation; electrically noisy industrial supplies |
| VI — line-interactive | Regulates voltage from the mains and transfers to the inverter on failure | Office IT where a brief transfer is acceptable |
| VFD — passive standby | Passes the mains through and transfers on failure | Individual desktops; not process loads |
- Static bypass. Transfers the load to the bypass supply on overload or internal fault. A UPS left in bypass after an event protects nothing, so make bypass status a monitored alarm.
- Maintenance bypass. An external wrap-around bypass lets the UPS be isolated for service without shutting the load down. Specify it at the start; it is hard to add later.
- DG compatibility. On DG supply the frequency moves more than on the grid, and a UPS may refuse to synchronise its bypass and block transfers. The rectifier's input current distortion also loads the DG's alternator. Give the DG supplier the UPS input data (see DG expansion and synchronisation).
- Redundancy. Where a UPS fault would stop the process, two units in parallel or a modular UPS with an extra module (N+1) remove that single point of failure — provided the distribution downstream is designed so one fault does not take everything with it.
Battery autonomy: ageing, temperature and the OEM discharge tables
- Use the manufacturer's discharge tables. Runtime is read from the battery maker's constant-power tables (watts per cell or per block) to a stated end voltage at a stated temperature. A simple ampere-hour calculation overstates runtime at the high discharge rates a UPS draws.
- Size for end of life. Capacity falls with age. IEEE 1188, the recommended practice for maintaining, testing and replacing valve-regulated lead-acid batteries, treats capacity below 80% of the manufacturer's rating as a replacement point — so a bank that only just meets the autonomy when new will fall short long before it is replaced. Apply an ageing allowance and design for end-of-life capacity.
- Temperature works both ways. High room temperature shortens battery life (the manufacturer's life-versus-temperature curve shows by how much); low temperature reduces the capacity available. Air-condition the battery room and log its temperature.
- Chemistry. Valve-regulated lead-acid remains the lower-capital option. Lithium-ion (commonly lithium iron phosphate) costs more upfront, weighs less and generally tolerates temperature and cycling better, but brings its own fire-safety requirements (see lithium-ion UPS room fire safety). Compare both over the planned life, using each manufacturer's data.
Decision table: which UPS arrangement fits
| What the schedule shows | Arrangement to evaluate | Watch-outs |
|---|---|---|
| A few kW of controls spread across many panels | Small online units at the panels, or a central online UPS with dedicated distribution | Battery upkeep at many locations versus long distribution runs |
| Server or network room backed by a reliable DG | Central online UPS with minutes of autonomy, static and maintenance bypass | Measured DG start-and-transfer time; bypass synchronisation on DG |
| No DG, or a DG that may fail to start | Autonomy sized to a controlled shutdown of each process | Shutdown sequences must be automated and tested |
| A process drive that must ride through dips | The drive maker's ride-through option or a UPS on the control supply first; a full UPS only if needed | The drive manufacturer's guidance; UPS sized for the drive's input current |
| A load that cannot tolerate a UPS fault | Parallel or modular N+1 with maintenance bypass | Downstream distribution must not reintroduce a single point of failure |
| Office IT where a brief transfer is acceptable | Line-interactive units may be enough | Not for process control or instrumentation |
Testing and maintenance that prove the autonomy
- Trend battery health — impedance or conductance against the commissioning baseline — instead of relying on float voltage alone.
- Run a controlled transfer test and, at intervals, a discharge test, so the autonomy is demonstrated rather than assumed. These can be written into an AMC.
- Keep bypass status, alarms and battery-room temperature on a monitored point.
- New input and output feeders and distribution boards for a UPS are alterations to the electrical installation: carried out by a contractor licensed for the work by the State Government, and approved or self-certified before connection under the CEA (Measures relating to Safety and Electric Supply) Regulations, 2023 (Regulations 31 and 45).
For the supply side around a UPS — DG sets, transformers and distribution — see Secured Engineers' electrical infrastructure service, or request a scope review with the schedule above.
Sources and how to verify
- International Electrotechnical Commission, IEC 62040-3:2021, Uninterruptible power systems (UPS) — Part 3: Method of specifying the performance and test requirements.
- IEEE, IEEE 1188, Recommended Practice for Maintenance, Testing, and Replacement of Valve-Regulated Lead-Acid (VRLA) Batteries for Stationary Applications.
- Central Electricity Authority, CEA (Measures relating to Safety and Electric Supply) Regulations, 2023, notified 8 June 2023: Regulations 31 and 45.
Confirm locally: your DG's measured start-and-transfer time, the UPS and battery manufacturers' data for the equipment actually selected, the fire-safety requirements for the battery room, and the voltage your state has notified for self-certification.
Frequently asked
What load information do I need before sizing a UPS for a factory or office?
A list of the loads that must not lose power, with measured kW and kVA, their inrush behaviour, how long each must be carried and why, whether a DG is available and its measured start-and-transfer time, the redundancy the process needs, and the growth expected on the same UPS. Nameplate totals are a poor substitute for a few days of power-analyser readings.
Which site conditions could change the scope or budget of a UPS project?
Battery-room space, floor loading and air-conditioning; ventilation for lead-acid batteries and fire-safety requirements for lithium-ion; whether the DG can feed the UPS input without frequency or distortion problems; cable routes to loads spread across the plant; earthing; and whether the load can be shut down for the changeover or needs a maintenance bypass first.
What should I send an engineer for a UPS sizing review?
The single-line diagram, the equipment schedule marking the loads you want protected, measured kW where available, the autonomy each load needs and why, the DG rating and its measured start-and-transfer time, details of any existing UPS and batteries, and recent demand records. That is enough to propose an arrangement and check it against the DG.
How many minutes of battery does a UPS need if the site has a DG?
Enough to cover the DG's measured start-and-transfer time with margin, plus the time to shut processes down safely if the DG fails to start, calculated at the battery's end-of-life capacity. That is usually minutes rather than hours; battery beyond it buys protection the DG already provides.
Related services
More insights
Busbar or Cable? The Rising-Mains Decision You Only Get to Make Once
Once the shafts are cast and the cables pulled, changing your mind costs a shutdown. Here is the honest engineering comparison between busbar trunking and cable risers for Indian industrial and commercial buildings — including where each one genuinely wins.
ElectricalElectrical Safety Audit Before a Renovation or Expansion
An uneventful electrical system is not a verified one. Before a renovation or expansion adds load, document the installation, have competent people test it, and turn the findings into an evidence-ranked rectification scope — the additions themselves need inspector approval or self-certification before they are connected.
More on Electrical & Power · all insights