UPS Sizing & Runtime Budget Guide
A UPS that is undersized, wrongly configured, or fitted with batteries that cannot survive the room they sit in will fail exactly when it is needed. This guide sets out how to size, specify and budget UPS protection for Bahrain and Saudi Arabia facilities.
In short: size a UPS from the real watts and power factor of the connected load with headroom for growth, choose online double-conversion for any load that cannot tolerate a transfer, set runtime against a documented generator handover or ride-through requirement rather than a round number, and confirm battery chemistry and room ventilation together — because Gulf ambient temperature is one of the biggest drivers of battery life and unplanned replacement cost.
This guide is for facilities, IT and electrical decision-makers specifying UPS protection for a comms room, data centre, control room or production line in Bahrain or Saudi Arabia — whether for a new installation, an expansion, or a replacement of ageing equipment reaching the end of its service life.
What this guide covers
UPS specification touches electrical design, IT criticality, battery engineering and facilities planning at the same time, which is why it is so often under-specified. This guide covers the practical decisions in order: assessing the real load, choosing a topology, deciding phase configuration, setting runtime against generator handover, selecting battery chemistry, deciding redundancy, and planning bypass, monitoring, environment and lifecycle together as one specification rather than as separate purchases.
Load assessment: VA vs watts and power factor
A UPS is rated in both kVA (apparent power) and kW (real power). The relationship between the two is the power factor of the connected equipment. Modern IT and server loads typically have a power factor close to unity, but mixed loads — motors, older equipment, some industrial control gear — can have a power factor well below 1.0, meaning the UPS needs more apparent power capacity than the real wattage alone suggests.
Sizing from nameplate VA ratings added together almost always overstates the real load, because nameplate figures are worst-case maximums, not typical draw. The safer approach is to measure actual running load where equipment already exists, or calculate it carefully from manufacturer power consumption data for new equipment, then add headroom — commonly in the order of 20–30% — for growth, inrush current at power-up, and future equipment additions, rather than sizing to the exact load on day one.
Topology: offline, line-interactive, online double conversion
| Topology | Best suited to | Considerations |
|---|---|---|
| Offline / standby | Individual workstations, low-criticality loads | Brief transfer time on switch to battery; lowest cost |
| Line-interactive | Small server rooms, network cabinets, retail back-office | Regulates voltage without switching to battery for minor sags; moderate cost |
| Online double conversion | Data centres, comms rooms, control rooms, any zero-transfer-time requirement | Output continuously regenerated; highest protection, highest running cost due to conversion losses |
The decision comes down to what the load can tolerate. If a few milliseconds of transfer time, or a brief dip in voltage regulation, would cause equipment to reset or drop connections, online double-conversion is the appropriate choice regardless of load size. Where the load is genuinely tolerant — a single workstation, a small retail point-of-sale system — a lower topology reduces cost without a meaningful increase in risk.
Single-phase vs three-phase
Small UPS units up to a few kVA are normally single-phase and suit comms cabinets, small server rooms and individual racks. Larger loads — data halls, plant rooms, buildings with substantial three-phase distribution already in place — are usually served by three-phase UPS systems, which balance load across the incoming supply and scale more efficiently at higher capacity. The decision should follow the site's existing electrical distribution and the total protected load, confirmed with the site's electrical design rather than assumed from the UPS capacity alone.
Runtime vs generator handover
Runtime should never be an arbitrary number. It is sized against one of two outcomes: riding through short utility interruptions without any other backup, or bridging the gap until a standby generator starts, transfers and stabilises. Where a generator is present, the UPS runtime needs to comfortably exceed the generator's start and transfer time, plus a margin for a failed first start attempt — a generator that takes 30–60 seconds to transfer should not be paired with a UPS sized for exactly that duration.
Where there is no generator, runtime is set against how long the site is prepared to operate on battery before a controlled or automatic shutdown, which should be documented and tested rather than assumed.
Battery chemistry: VRLA vs lithium
| Factor | VRLA (sealed lead-acid) | Lithium (LFP) |
|---|---|---|
| Upfront cost | Lower | Higher |
| Footprint | Larger for equivalent runtime | Significantly smaller and lighter |
| Temperature sensitivity | Life reduces markedly at elevated ambient temperature | Wider tolerance, generally more stable in warm rooms |
| Typical replacement cycle | Shorter, more predictable interval, especially in warm rooms | Longer expected service life |
| Monitoring needs | Regular internal resistance and float voltage checks | Battery management system provides ongoing cell-level data |
In Gulf ambient conditions, the practical difference between the two chemistries is often decided by the battery room. A well-ventilated, temperature-controlled room narrows the gap; a room with poor ventilation or no cooling will shorten VRLA life significantly and make the lithium premium easier to justify over the full lifecycle. Footprint also matters where floor space in a comms room or plant room is constrained.
Redundancy: N+1, parallel, dual feed
Redundancy is a decision about acceptable risk, not a default specification. Options include:
- N+1 — one additional UPS module beyond what the load requires, so a single failure or scheduled maintenance does not remove protection.
- Parallel systems — multiple UPS units sharing the load, providing both capacity and redundancy, common where load growth is expected.
- Dual feed / dual bus — critical equipment fed from two independent UPS systems, used where the connected equipment itself has dual power inputs.
The right level of redundancy is set by the cost of downtime to the specific load, not by a general preference for resilience. A single-module UPS with a good maintenance contract may be entirely appropriate for a lower-criticality load.
Bypass and maintenance bypass
An automatic bypass allows the load to be transferred to raw utility supply if the UPS itself detects an internal fault, so a UPS failure does not necessarily mean the load loses power. A maintenance bypass is a manually operated switch that allows the UPS to be isolated completely — for testing, battery replacement, or eventual replacement of the unit — without interrupting the connected load. Sites that omit a maintenance bypass at installation typically discover the gap the first time the UPS needs servicing, when the only options are an accepted outage or an improvised temporary bypass.
Harmonics and input filtering
UPS input rectifiers and other non-linear loads on the same electrical system can introduce harmonic distortion, which affects power quality, generator sizing and, in some cases, neighbouring equipment. Larger three-phase UPS systems commonly specify input filtering or higher-pulse rectification to limit harmonics fed back into the site's electrical supply. This is a design detail that should be confirmed with the electrical consultant on larger installations, particularly where a standby generator is already sized close to its limit.
Environment and ventilation in Gulf ambient conditions
UPS units and batteries generate heat and are rated for a specific operating temperature range. In Bahrain and Saudi Arabia, where ambient conditions can be extreme for parts of the year, the room housing the UPS and its batteries needs dedicated cooling sized for the equipment's actual heat output — not simply the general room air conditioning. Battery rooms in particular benefit from stable, cooler temperatures, since elevated heat is one of the most significant factors shortening VRLA battery life. Ventilation also matters for any battery technology that can release gas under fault conditions, and the room layout should follow the manufacturer's clearance and ventilation guidance.
Monitoring and shutdown software
A UPS should report its status — load level, battery health, runtime remaining and alarm conditions — to the people responsible for the site, not just display a light on the front panel. Network management cards, building management system integration, and shutdown agents installed on servers to trigger a controlled shutdown before batteries are exhausted are all part of a complete specification, not optional extras. Without shutdown software, an extended outage beyond the runtime simply ends in an uncontrolled loss of power to the connected equipment.
What drives cost
UPS project cost is driven by a combination of factors rather than the UPS unit price alone:
- protected load in kVA/kW and the topology required to serve it;
- runtime requirement and the resulting battery bank size;
- battery chemistry selected;
- redundancy level — N+1, parallel or dual feed;
- switchgear, distribution boards and maintenance bypass arrangements;
- civil and electrical works — battery room preparation, cooling, cable containment and earthing;
- monitoring integration and commissioning scope.
Two UPS systems of the same kVA rating can differ substantially in installed cost once runtime, redundancy and room preparation are accounted for, which is why a budgetary figure should always be tied to a specific load, runtime and redundancy assumption.
Common mistakes
- sizing from nameplate VA ratings rather than measured or calculated real load;
- no headroom for growth, so the UPS is already at capacity at handover;
- ignoring battery room temperature and ventilation when selecting chemistry;
- no runtime verification test, so the assumed autonomy is never actually confirmed;
- omitting a maintenance bypass, forcing an outage for the first service visit;
- redundancy decided by budget rather than by the criticality of the load;
- no shutdown software, leaving an extended outage to end uncontrolled;
- battery replacement left unbudgeted until the batteries are already failing.
Commissioning and acceptance criteria
- Step 01Load bank test
The UPS is tested under a controlled resistive load to verify it holds rated output and voltage regulation as specified.
- Step 02Runtime verification
The system is run on battery under representative load to confirm the calculated runtime is actually achieved, not just modelled.
- Step 03Transfer and bypass test
Automatic and maintenance bypass transfers are exercised to confirm the load transfers cleanly in both directions.
- Step 04Alarm and monitoring test
Alarm outputs, network cards and shutdown agents are verified end to end, including notification to the responsible team.
Lifecycle and battery replacement planning
A UPS is a long-life asset, but its batteries are a consumable with a shorter, environment-dependent life. Planning for replacement means recording the installation date, chemistry and expected life against room conditions, scheduling periodic testing — including internal resistance or impedance checks rather than relying on voltage alone — and budgeting for replacement before failure rather than reacting to it. This is typically covered under an annual maintenance contract, which keeps testing on schedule and gives early warning of degrading cells.
Decision checklist
- Real load in watts and kVA measured or calculated, not assumed from nameplates
- Headroom allowed for growth and inrush
- Topology matched to the load's tolerance for transfer time
- Single-phase or three-phase confirmed against site distribution
- Runtime set against a documented ride-through or generator handover time
- Battery chemistry selected with room temperature and footprint considered
- Redundancy level agreed against load criticality
- Maintenance bypass included in the design
- Harmonics and input filtering reviewed for larger three-phase systems
- Battery room ventilation and cooling sized for the equipment
- Monitoring, network cards and shutdown software specified
- Load bank, runtime and alarm tests included in the acceptance plan
- Battery replacement budget and schedule agreed at handover
Frequently asked questions
Reviewed by the MTT Engineering Team · Published 30 August 2026
Related MTT capabilities
For a related building block of a resilient facility, see our server room and data centre planning guide, our industrial and manufacturing solutions, or browse our case studies. For a documented proposal, get in touch through contact MTT.
Plan your project with MTT
Tell us about your load, criticality and existing electrical infrastructure and an MTT engineer will arrange a survey and a documented proposal.