Online UPS Systems for Servers & Critical Infrastructure: Complete Guide India

A server room in India faces power threats that simply don't exist at the same frequency in Europe or North America: voltage that swings from 160V to 280V within the same hour, harmonic distortion from adjacent industrial loads, micro-outages lasting 80–400ms that corrupt running database transactions, and grid frequency that routinely drops to 49.3 Hz during peak demand. For this environment, the power protection device matters enormously — and the wrong choice creates risk that negates the investment in the servers themselves.

This guide covers online UPS systems for India's server rooms, data centres, hospitals, and critical infrastructure — from fundamental topology comparison to data centre tier specifications, harmonic distortion standards, modular UPS architecture, battery technology selection, and everything a facility manager or IT infrastructure owner needs to make a correct, defensible purchasing decision.

The Three UPS Topologies: Why Only One Is Right for Critical Infrastructure

The UPS market in India offers three distinct protection topologies at radically different price points. Understanding the technical differences prevents the expensive mistake of buying a cheaper topology that fails to protect under India's grid conditions:

Topology	How Load Is Powered	Transfer Time on Failure	Voltage Regulation	Frequency Regulation	Harmonic Isolation
Offline / Standby	Directly from grid; battery only on failure	10–25 ms	None (pass-through)	None	None
Line-Interactive	Grid through AVR transformer; battery on failure	4–8 ms	±8–12% (AVR range)	None (pass-through)	Partial
Online Double-Conversion	Always through rectifier + inverter; battery on DC bus	0 ms (zero transfer)	±1% continuous	±0.05% (crystal-controlled)	Complete isolation

The transfer time difference is the decisive factor for servers. A server's SMPS (Switch Mode Power Supply) can ride through a 4–8ms gap — most IT equipment is specced to handle up to 10ms. But India's line-interactive UPS units in budget categories often take 10–15ms in practice. More critically, during a voltage sag (not a full outage) that lasts 200ms, a line-interactive UPS may not switch to battery at all — it attempts to regulate the sag through the AVR transformer, but if the sag is below the AVR's floor, the load receives undervoltage for the full 200ms. Running database transactions during that 200ms see power supply dropouts that corrupt in-flight writes.

An online double-conversion UPS eliminates all of this. The load is always powered by the inverter. The inverter's output is synthesised from a crystal-controlled oscillator — 50.00 Hz, 230V ±1%, regardless of what the grid input looks like. Grid events (sags, swells, outages, frequency drift, harmonics) are completely invisible to the load.

Double-Conversion vs Delta-Conversion: The Other Online Topology

Double-conversion is not the only online topology. Delta-conversion (also called delta online) is an alternative used by some high-efficiency enterprise UPS systems:

• Double-conversion: All load current passes through rectifier → DC bus → inverter. Complete power reconditioning. Efficiency: 92–96% (with ECO mode bypass available for higher efficiency when grid is clean).
• Delta-conversion: A small delta transformer handles only the power difference between input and output; most power flows directly. More efficient (96–98%) but provides less isolation — not true double-conversion. Used in Eaton 9395 and similar enterprise-grade systems.

For Indian conditions with poor power quality, true double-conversion is preferred — delta-conversion's partial power flow means grid disturbances still partially reach the load. Delta-conversion is appropriate in high-quality grid environments (Tier III/IV data centres with redundant utility feeds) where efficiency gain justifies the reduced isolation.

ECO Mode: When to Use It and When Not To

Most modern online UPS systems include an ECO (Economy) mode that operates line-interactive when input power is within acceptable limits, switching to double-conversion only when needed. This raises efficiency from ~94% to ~98% — a significant energy saving in large installations.

ECO mode is appropriate when:

• Your facility has a stable, high-quality power supply (Tier III/IV data centre with dedicated utility feed)
• The UPS is protecting IT equipment with wide-range SMPS that tolerate brief gaps
• Energy cost savings are a priority and occasional 4–8ms transfers are acceptable

ECO mode is NOT appropriate when:

• Your grid has frequent voltage sags, swells, or harmonic distortion (most Indian commercial areas)
• The UPS is protecting medical equipment, industrial automation, or any load that cannot tolerate brief interruptions
• Your input voltage frequently approaches the ECO mode limits (typically ±10% of nominal)

For most Indian server rooms outside Grade A office parks with dedicated power infrastructure, keep ECO mode disabled and accept the ~2–4% efficiency penalty in exchange for full isolation.

Why Indian Grid Conditions Demand Online Double-Conversion

India's power quality problems are well-documented and consistent. A power quality audit at any Indian commercial installation will typically reveal:

Power Quality Problem	Typical Indian Grid	IEC 61000-4 Spec for IT Equipment	Offline/LI UPS Protection	Online UPS Protection
Voltage sags (below 207V)	5–15 events/day in commercial areas	Must ride through ≤10ms at 0V	Partial — AVR only corrects ±15%	Complete — load never sees sag
Voltage swells (above 253V)	2–8 events/day	Must withstand 250V continuous	Partial — AVR clips swell	Complete
Micro-outages (50–500ms)	1–5 per day in many cities	Must ride through ≤10ms	Fails — 10–25ms transfer gap	Complete — zero gap
Harmonic distortion (THDU)	8–15% THD in industrial areas	Equipment may specify ≤5% THD	None — pass-through	Complete — inverter output ≤3% THD
Frequency deviation (±0.5Hz)	49.3–50.7 Hz typical range	47–53 Hz tolerance (wide)	None — pass-through	Complete — crystal-controlled 50.00 Hz
Electrical noise / spikes	Common near industrial loads	IEC 61000-4-5 surge immunity	Partial (MOV protection only)	Complete — DC bus blocks all spikes

Data Centre Tier Classification and UPS Requirements

The Uptime Institute's Tier classification system defines four levels of data centre reliability. Each tier has specific UPS architecture requirements:

Tier I — Basic Capacity

Single path for power and cooling. No redundancy. 99.671% uptime (28.8 hours downtime/year). UPS requirement: single online UPS per power distribution unit, sized for full load. Acceptable for small server rooms and SME data closets where occasional outages are tolerable.

Tier II — Redundant Components

Single path with redundant components (N+1 batteries, dual PDUs). 99.741% uptime. UPS requirement: N+1 battery modules and redundant power feeds from UPS to servers. Appropriate for SME server rooms and small colocation facilities.

Tier III — Concurrently Maintainable

Multiple power and cooling paths, but only one active at a time. 99.982% uptime (1.6 hours downtime/year). UPS requirement: 2N or N+1 UPS configuration with static transfer switches; ability to maintain any UPS component without load impact. Standard for enterprise data centres and colocation facilities in India.

Tier IV — Fault Tolerant

Multiple active power and cooling paths. 99.995% uptime (26 minutes downtime/year). UPS requirement: 2N UPS configuration (two independent UPS systems each carrying full load), dual-corded servers (servers with two PSUs connected to independent UPS), automatic fault detection and isolation. Required for financial institutions, mission-critical healthcare, and national infrastructure.

Most Indian corporate server rooms and smaller data centres target Tier II–III. For Tier II, a single quality online UPS with N+1 battery modules and bypass capability is sufficient. For Tier III, plan for modular UPS architecture with parallel redundancy.

Modular UPS Architecture: Scalability and Redundancy

Traditional monolithic online UPS (one unit, fixed capacity) has given way to modular UPS in enterprise applications. Understanding the difference helps with long-term infrastructure planning:

Traditional Monolithic UPS

• Fixed capacity — 20 kVA, 40 kVA, 80 kVA units purchased as a block
• If the single unit fails, all load is lost (unless a bypass is active)
• Sized for future capacity at purchase — pays for capacity years before it's needed
• Servicing requires load transfer to bypass — brief exposure to grid power

Modular Online UPS

• Chassis holds N power modules (4–20 kW each), any of which can be hot-swapped
• Capacity added by inserting additional modules — no system downtime
• Redundancy achieved by having one more module than needed (N+1) — if one fails, others carry the load
• Failed module hot-swapped in 5 minutes without load impact
• Pay for capacity as you grow — significant CAPEX savings for expanding facilities

Modular UPS is the correct choice for: rapidly growing server rooms, facilities that need Tier III concurrently maintainable operation, and any installation above 40 kVA where monolithic redundancy would require two full 40 kVA UPS units (expensive and space-consuming).

Su-vastika's modular online UPS range starts at 20 kVA chassis with 5 kW modules, expandable to 80 kVA in the same chassis — with N+1 redundancy configurable from initial installation.

Harmonic Distortion: The Hidden Power Quality Threat

Online UPS systems have two harmonic specifications that are frequently misunderstood during procurement:

Output THDU (Total Harmonic Distortion — Voltage)

The harmonic distortion on the UPS output voltage — what the servers actually receive. Should be ≤3% at all load conditions. A UPS with high output THDU causes premature PSU failure in servers, overheating in motor loads, and data corruption in precision instruments. All quality online UPS specify ≤3% output THDU — verify this in the spec sheet, not just the sales brochure.

Input THDI (Total Harmonic Distortion — Current)

The harmonic current drawn by the UPS from the grid. This matters because high input THDI contaminates the facility's electrical system, affecting other equipment and causing neutral conductor overheating (a fire risk in multi-floor buildings).

Legacy online UPS systems with 6-pulse or 12-pulse rectifiers draw high input THDI (25–30% for 6-pulse). Modern UPS with IGBT active rectifiers achieve input THDI ≤3% — essentially unity power factor with clean sinusoidal current draw. This is now the standard for any UPS above 20 kVA in India's energy-conscious market.

Specifying ≤5% input THDI is the minimum acceptable standard for new UPS procurement. Older 6-pulse or 12-pulse rectifier UPS systems should be identified for upgrade when battery replacement is due.

Input Power Factor

Closely related to input THDI: modern online UPS with IGBT active rectifiers achieve input power factor of 0.99 or better — meaning almost all power drawn from the grid is real power, not reactive. This reduces your building's reactive power demand, potentially avoiding power factor penalties from DISCOM (which apply when facility PF drops below 0.9 in most Indian tariff structures).

Hospital and Medical Facility UPS: Additional Requirements

Medical facilities in India face additional power quality and safety requirements beyond standard commercial UPS specifications:

• IS 13947 (Medical Electrical Equipment): Medical equipment in patient care areas must be powered through isolated power supplies (IT systems — Isolated Terra) or UPS with galvanic isolation. Standard UPS output is typically TN-S (earthed neutral) — verify isolation transformer availability for medical applications.
• IEC 60601-1 (International standard adopted in India): Medical equipment power supplies must meet leakage current limits ≤0.5mA for equipment touching patients. The UPS system must not introduce additional earth leakage current.
• Essential power systems (NBC 2016 Part 4): Hospitals must maintain three levels of emergency power: critical branch (life support, operating theatres — generator + online UPS), equipment branch (imaging, lab), and lighting branch. UPS backs up the critical branch; generator covers all three.
• Zero transfer time: Life-support equipment (ventilators, infusion pumps, cardiac monitors) requires zero transfer time UPS. Online double-conversion is mandatory — not optional — for ICU and operating theatre power systems.
• Battery runtime: Hospital UPS should provide minimum 30 minutes runtime at full load — enough for generator startup, stabilisation, and ATS (Automatic Transfer Switch) operation even if the generator has startup issues.

How to Size an Online UPS: Complete Process

Step 1: Audit Actual Power Consumption

Nameplate power ratings on servers and network equipment are worst-case maximums — actual consumption is typically 40–60% of nameplate. Measure with a power meter (Fluke or equivalent) or check the server's ILO/IPMI power reading. Using nameplate ratings results in 2–3x oversizing and wasted capital.

Step 2: Calculate Total kVA Load

Equipment Type	Typical Actual Consumption	Power Factor	kVA
1U Rack Server (mid-range)	150–300W	0.95–0.99	155–315 VA
Blade server chassis (10 blades)	2,000–4,000W	0.95–0.99	2.1–4.2 kVA
Network switch (48-port)	150–400W	0.95	160–420 VA
Firewall / router	100–300W	0.95	105–315 VA
SAN / NAS storage (per unit)	300–800W	0.95	315–840 VA
KVM switch + console	30–80W	0.95	32–84 VA

Step 3: Apply 75% Loading Rule

Never load an online UPS above 75% of rated capacity. Reasons: (1) load fluctuations — servers spike to 120–150% of average during backup operations or batch jobs; (2) future growth — plan for 20–30% capacity addition over 3 years; (3) efficiency — most online UPS reach peak efficiency at 70–80% load.

Required UPS kVA = Total load kVA ÷ 0.75

Step 4: Determine Battery Runtime

Runtime planning should address two scenarios:

• Short outage scenario (most common): 5–15 minutes runtime to ride through typical Indian grid outage and allow graceful shutdown if power doesn't return. Standard internal batteries on most UPS cover this.
• Extended backup scenario: If you need 30–120 minutes runtime (e.g., generator not available, hospital application, trading floor that cannot lose connectivity), specify external battery cabinets (EBC) or select a UPS model with EBC expansion ports.

Runtime calculation: Runtime (min) = (Battery Wh × 0.95 efficiency) ÷ Load W × 60

Example: 20 kVA UPS at 14 kW load with 20 kWh external battery: (20,000 × 0.95) ÷ 14,000 × 60 = 81 minutes

Step 5: Redundancy Configuration

Configuration	Setup	Protection Level	Recommended For
N (single unit)	One UPS sized for full load	No UPS redundancy	Non-critical servers, small offices
N+1 modular	One extra power module in modular UPS chassis	Single module failure tolerated	SME server rooms, Tier II
N+1 parallel	Two UPS units in parallel, each sized for full load	Single UPS failure tolerated	Enterprise, Tier III
2N	Two independent UPS systems, each carrying full load; dual-corded servers	Complete UPS system failure tolerated	Tier IV, financial, healthcare

Battery Technology Selection: VRLA vs LiFePO4 for Online UPS

The battery decision for an online UPS is higher-stakes than for a home inverter — larger battery banks, more critical applications, and machine room environments with restricted access for maintenance.

VRLA (Valve Regulated Lead-Acid)

Still the dominant technology for UPS batteries globally, but with significant limitations in Indian conditions:

• Rated at 5 years (float service life at 20°C). At Indian machine room temperatures of 35–40°C, actual life is 2.5–3.5 years — every 10°C above 20°C halves VRLA battery life
• Replacement cost for a 40 kVA UPS battery bank: ₹2–4 lakh every 3 years
• Heavy: a 40 kVA UPS battery bank weighs 400–800 kg — floor load rating must be verified
• Reliable, proven, widely available, lower upfront cost

LiFePO4 (Lithium Iron Phosphate)

Premium option with compelling TCO advantages for larger installations:

• 10–15 year service life at 25°C; 8–10 years at 35–40°C — significantly longer than VRLA in Indian conditions
• 60–70% lighter: a 40 kVA LFP battery bank weighs 120–200 kg vs 400–800 kg for VRLA — enables installations where VRLA floor loading would fail
• Recharges in 1–2 hours vs 6–10 hours for VRLA — critical in high-outage areas where multiple daily outages must not deplete the battery
• Higher upfront cost: LFP battery module costs 2–3x VRLA of equivalent energy capacity
• TCO crossover: LFP typically reaches TCO parity with VRLA at year 5–7 through avoided battery replacements

10-Year Battery TCO Comparison (40 kVA UPS, Indian Conditions)

Item	VRLA	LiFePO4
Initial battery cost	₹3,00,000	₹7,00,000
Year 3 replacement (VRLA only)	₹3,50,000	—
Year 6 replacement (VRLA only)	₹4,00,000	—
Year 9 replacement (VRLA only)	₹4,50,000	—
Battery disposal (VRLA hazardous waste, 3x)	₹30,000	₹5,000
Downtime cost (battery failure, 2 incidents)	₹2,00,000 (estimated)	₹20,000 (BMS warning → planned replacement)
10-year battery TCO	₹17,30,000	₹7,25,000

For any UPS installation above 20 kVA in a high-availability application, LiFePO4 delivers substantial TCO savings over 10 years despite the higher upfront cost — and eliminates the reliability risk of end-of-life VRLA batteries failing between planned replacement cycles.

Key Specifications Checklist: What to Verify Before Buying

Specification	Minimum Acceptable	Best Practice
Topology	Online double-conversion	Online double-conversion with IGBT rectifier
Input voltage range	160–280V	110–300V
Output voltage regulation	±3%	±1%
Output frequency regulation	±0.5 Hz	±0.1 Hz (crystal-controlled)
Output THDU	≤5%	≤3%
Input THDI	≤10%	≤3% (IGBT active rectifier)
Input power factor	≥0.95	≥0.99
Efficiency at 50% load	≥90%	≥94%
Efficiency at 75–100% load	≥92%	≥96%
Transfer time (grid failure)	0 ms (online topology)	0 ms
Transfer time (to bypass)	≤4 ms	≤2 ms
MTBF	≥100,000 hours	≥150,000 hours
Battery monitoring	Pack voltage and alarm	Cell-level BMS with predictive failure alert
Communications	RS232 or USB	SNMP card, Modbus RTU, dry contacts
Bypass	Manual maintenance bypass	Static bypass + maintenance bypass
Certification	BIS (mandatory)	BIS + CE + IEC 62040-1

Brand Comparison: Major Online UPS Suppliers for Indian Market

Brand	Range (single-phase)	Range (3-phase)	LFP Option	Modular	India Service
Su-vastika	1–20 kVA	10–500 kVA	Yes	Yes (20–80 kVA)	150+ cities
APC by Schneider	1–20 kVA	10–500 kVA	Partial (select models)	Yes (Galaxy range)	Pan-India
Eaton	1–20 kVA	10–1600 kVA	Yes (9SX Li)	Yes (9PX modular)	Pan-India
Consul Neowatt	1–20 kVA	10–500 kVA	Partial	Limited	Good coverage
Legrand / Numeric	1–20 kVA	10–200 kVA	No	Limited	Wide distributor network
Vertiv (Liebert)	1–20 kVA	10–500 kVA	Yes	Yes	Metro-focused
Delta Electronics	1–20 kVA	10–800 kVA	Yes	Yes	Growing in India

Key differentiator for Su-vastika: wide Indian service network (150+ cities) vs global brands whose service SLAs outside metros can be 24–72 hours — critical for Tier II/III uptime requirements.

Remote Monitoring and Management: The Non-Negotiable Feature

Every online UPS protecting a server room must have network management capability. The reason: UPS battery failures are the leading cause of data centre downtime — and they're almost always predictable with monitoring.

Required monitoring capabilities:

• SNMP network management card: Integrates with existing NMS (Nagios, Zabbix, Prometheus) for centralised alerting. Should support SNMPv3 (authentication and encryption — SNMPv1/v2c are insecure).
• Automatic server shutdown: NUT (Network UPS Tools) or vendor software monitors battery runtime and triggers graceful OS shutdown before power runs out. Protects data integrity — no "hard crash" when battery dies.
• Battery impedance trending: Advanced BMS measures battery internal resistance at each charge cycle. Rising impedance is the leading indicator of impending battery failure — detectable 3–6 months before failure. This converts unplanned outages (battery fails during actual power cut) into planned maintenance (battery replaced on schedule).
• Event logging: Timestamped records of all power events (sags, outages, switchovers, overloads) — essential for post-incident root cause analysis and insurance claims.
• Environmental sensors: Temperature and humidity in the UPS enclosure — alert when machine room cooling fails before the UPS battery is damaged.

Installation Requirements and Best Practices

• Temperature: Online UPS electronics are rated for 0–40°C operating range. Battery life (VRLA) degrades rapidly above 25°C — every 10°C increase halves battery life. Server room cooling should maintain 22–24°C for optimal battery longevity. If cooling fails regularly, specify LFP batteries.
• Input power feed: For UPS above 10 kVA, install a dedicated circuit from the main LT panel — not shared with HVAC or other large loads. Shared circuits cause voltage sags when compressors start, unnecessarily switching the UPS to battery.
• Bypass panel: Install a manual bypass switch (MBS) upstream of the UPS — allows UPS to be isolated for maintenance while servers operate on raw grid power. Required for Tier II and above (concurrently maintainable).
• Cable sizing: Size input and output cables for 125% of UPS rated current. Undersized cables cause voltage drop and excessive heat. For a 40 kVA three-phase UPS at 58A per phase: minimum cable size is 16 mm² copper.
• Earthing: UPS output neutral must be earthed per IS 3043. Floating neutral UPS output causes common-mode voltage issues that corrupt data in SCSI/SAS storage and damage certain networking equipment.
• Ventilation clearance: Maintain minimum 600mm rear clearance and 300mm side clearance for heat dissipation. A 40 kVA online UPS dissipates 1.6–2.5 kW as heat — this must be accounted for in room cooling calculations.

Annual Maintenance Schedule

Frequency	Task
Monthly	Check SNMP dashboard — no alarms; verify battery float voltage within spec; inspect ventilation fans operational
Quarterly	Battery impedance test (via BMS or external tester); connection torque check on battery terminals; filter cleaning (dusty Indian environments clog UPS cooling vents rapidly)
Bi-annually	Full battery discharge test under load — verify actual runtime matches specification; calibrate battery monitoring after test
Annually	Complete UPS preventive maintenance by certified engineer: capacitor inspection, IGBT thermal imaging, bypass switch operation test, static transfer switch test, full performance verification
As triggered by monitoring	Battery replacement when impedance rises 20%+ above baseline, or capacity test shows ≤80% of rated — do not wait for failure

Frequently Asked Questions

What is the difference between a UPS and an inverter for servers?

A home inverter (offline topology) switches to battery in 10–25ms when grid fails — acceptable for home appliances, not for servers. An online UPS (double-conversion) has zero transfer time — the load is always on the inverter output. For servers, the minimum acceptable UPS topology is line-interactive (4–8ms transfer). For databases, financial systems, or any application where transaction integrity matters, online double-conversion is mandatory.

How do I know if my server room UPS needs replacement vs battery replacement?

The UPS electronics (inverter, rectifier, control board) typically last 10–15 years with proper maintenance. Batteries last 3–5 years (VRLA) or 8–12 years (LFP). Most UPS downtime events are battery failures, not UPS failures. Have the batteries impedance-tested annually. If the UPS is over 10 years old or showing efficiency drops, evaluate full unit replacement — older UPS with 6-pulse rectifiers waste 8–12% more power than modern IGBT units.

Can I use my existing generator with an online UPS?

Yes — but the generator must have good voltage and frequency stability to prevent the UPS from rejecting the generator input. Online UPS with wide input voltage range (160–280V) and frequency range (45–55 Hz) will accept most diesel generator output. If the UPS rejects generator input, check: generator voltage output (should be 230V ±5%), generator frequency (should be 50 ±1 Hz), and whether the generator's AVR is stable. Some older generators have high output THDU (harmonic distortion) that causes UPS input issues.

What runtime do I actually need for my server room UPS?

For most Indian server rooms: 15–20 minutes is sufficient for the most common scenario (short outage — grid returns before UPS depletes). If grid doesn't return within 15 minutes, it's likely a longer outage — the UPS should initiate graceful server shutdown rather than wait for battery to die. For facilities with generators, UPS runtime only needs to cover generator startup and ATS switchover (typically 30–60 seconds) plus a safety margin — 5–10 minutes total is adequate.

Is a single-phase UPS adequate for my server rack?

Single-phase UPS (up to 20 kVA) is adequate for small to medium server rooms (5–30 servers depending on density). Above 20 kVA, or whenever servers have dual PSUs that benefit from separate phase feeds, switch to three-phase UPS with three-phase output distribution. For high-density blade server environments, three-phase is typically required from 10 kVA upward to balance phase loading.

What is a static bypass and why does it matter?

A static bypass is a solid-state (IGBT or SCR) switch that transfers the load directly to grid power in under 4ms — used when the UPS overloads or fails. This is different from the manual maintenance bypass (which requires human intervention). The static bypass engages automatically and silently when: UPS is overloaded beyond its rated surge capacity, UPS internal fault is detected, or the UPS is shutting down. For most server loads, a 4ms static bypass transfer is survivable. Without a static bypass, a UPS failure means complete load loss.

Do I need BIS certification for an online UPS in India?

Yes. Online UPS systems are covered under the Electronics and IT Goods (Compulsory Registration) Order. BIS registration (IS 16242 for UPS systems) is mandatory for sale in India. Additionally, for installation in regulated facilities (hospitals, airports, financial institutions), CE marking and IEC 62040-1 compliance may be required by institutional IT security or facilities policies. Always verify BIS registration certificate number on the UPS or supplier documentation.

Conclusion: Online Double-Conversion Is the Only Correct Choice for Indian Critical Infrastructure

India's grid power quality — voltage sags, harmonic distortion, micro-outages, frequency drift — creates a demanding environment that only online double-conversion UPS handles completely. For any application where a power event would cause data loss, transaction corruption, equipment damage, or patient safety risk, the investment in online UPS is not discretionary.

The purchase decision then becomes: how much capacity, what redundancy level, VRLA or LFP batteries, and whether modular architecture is appropriate for your growth trajectory. For most Indian server rooms below 40 kVA, a quality traditional online UPS with IGBT active rectifier and LFP battery option is the right specification. Above 40 kVA, modular architecture with N+1 redundancy and LFP batteries delivers the best combination of reliability, maintainability, and 10-year TCO.

For buildings also running lift systems, see our Lift Inverter ERD guide. For commercial-scale energy storage beyond individual UPS systems, see our BESS Energy Storage guide.

Su-vastika's online UPS range covers 1 kVA single-phase to 500 kVA three-phase — with IGBT active rectifiers (≤3% input THDI), LiFePO4 battery options, modular architecture up to 80 kVA, SNMP network management, BIS certification, and a 150+ city service network with 4-hour SLA in metro areas.

Explore Su-vastika Online UPS Systems — 1 kVA to 500 kVA →