Data Center Panels

Data center electrical panel design is driven by availability architecture, not by panel ratings. The design starts from a single question — how power must continue to reach the critical IT load while any one piece of equipment is being maintained or has failed — and the panel's rated current, Form of separation, breaker type and fault level are then chosen to support that answer. Two independent A and B power paths, a defined bus-coupler and ATS philosophy, protection selectivity, and the removal of shared auxiliary components matter more than any single headline rating.

The consequence is that panel construction features and site availability classifications are separate things and cannot be substituted for one another. A Form 4 panel does not make a facility Tier IV, and a draw-out ACB does not by itself make a system concurrently maintainable. Both are useful, but availability is a property of the whole topology — including control supplies, auxiliary circuits and the protection study — not of one switchboard.

Why a Data Center Panel Is Different

A data center electrical panel has a very different job from an ordinary industrial distribution board.

In a typical plant, a maintenance shutdown may be planned around production.

In a data center, the electrical distribution system is often expected to allow equipment to be isolated, maintained or even fail without interrupting the critical IT load.

That requirement changes the way the complete electrical system—and therefore its PCC panels, ATS panels, bus couplers, UPS distribution boards and downstream switchboards—must be engineered.

The first question should not be:

“What size panel do we need?”

It should be:

“How must power continue to reach the critical load when equipment is maintained or fails?”

This guide explains the practical electrical-panel design considerations for modern data centers, including:

• A and B power paths
• N, N+1 and 2N redundancy
• PCC and main LV switchboards
• ATS and source-transfer philosophy
• Bus couplers
• Form of internal separation
• Fault-current ratings
• Breaker selectivity
• Neutral design
• UPS and harmonic loads
• Metering and monitoring
• FAT and maintainability

The objective is not to prescribe one universal architecture.

It is to understand the engineering questions that must be answered before a data center panel is manufactured.

Start With the Data Center Power Architecture

A data center panel cannot be designed correctly in isolation.

It is part of a complete power chain that may include:

Utility → HT Switchgear → Transformer → Main LT Switchboard → UPS → PDU / Busway → Rack

with additional sources such as:

DG Sets → Synchronization System → Emergency Bus

and in some facilities:

BESS / Energy Storage

The critical question is what happens when any one element in that chain is unavailable.

That is why data center electrical design begins with availability architecture, not just equipment ratings.

ISO/IEC 22237-3:2021 specifically addresses power distribution infrastructure for data centers as part of the wider ISO/IEC 22237 data-center facilities and infrastructure series.

What Do N, N+1 and 2N Mean?

These terms are often used when discussing data-center redundancy.

N

N means the infrastructure has enough capacity to support the required load.

For example:

If the critical requirement needs three UPS modules and exactly three are installed:

3 required + 0 spare = N

A failure or maintenance activity may reduce available capacity below the required level.

N+1

N+1 means the system has one additional capacity component beyond what is required.

Example:

If three UPS modules are required:

3 required + 1 redundant = N+1

One module can become unavailable while sufficient capacity remains—provided the rest of the architecture supports that redundancy.

2N

2N generally refers to two independent capacity paths, each capable of supporting the required load.

Conceptually:

Power Path A → Critical Load

and

Power Path B → Critical Load

Each path can be designed to support the required critical load independently.

This is substantially different from simply installing one spare breaker or one extra transformer.

Comparing the Three Redundancy Levels

The right-hand column is the one that causes disputes on site, because each level solves a narrower problem than people assume.

Concept	Capacity Installed	Can Something Be Taken Out of Service?	What It Still Does Not Guarantee
N	Exactly the capacity the critical load requires	No — removing any capacity component drops available capacity below the requirement	Any tolerance at all for maintenance or failure
N+1	Required capacity plus one redundant capacity component	One capacity component can be unavailable, provided the distribution path supports it	A second distribution path — a fault in the single path can still drop the load
2N	Two independent paths, each able to support the full critical load	Yes, an entire path can be isolated for maintenance	Genuine independence, if the two paths share transformers, control supplies or auxiliary circuits

Notice that N+1 is a statement about capacity components and 2N is a statement about distribution paths. A facility can have N+1 UPS modules and still have only one path to the load, which is why redundancy has to be specified for both capacity and distribution, not as a single label.

Tier III and Tier IV Are Not Panel Ratings

This is one of the most important distinctions in data-center projects.

A manufacturer may hear statements such as:

“We need a Tier III panel.”

or

“This must be a Tier IV Form 4 panel.”

Technically, the Uptime Institute Tier classification applies to the site infrastructure topology, not to an individual electrical panel.

Uptime Institute describes Tier III as Concurrently Maintainable and Tier IV as Fault Tolerant, with progressively more demanding requirements for capacity components and distribution paths.

Therefore:

A Form 4 panel does not automatically make a data center Tier IV.

And:

A draw-out ACB does not automatically make the system concurrently maintainable.

The complete electrical topology must support the intended availability objective.

This includes:

• Sources
• Transformers
• Switchgear
• UPS
• Distribution paths
• Control power
• Auxiliary systems
• Maintenance isolation
• Failure modes

Panel design must support that architecture.

A and B Power Paths

A common high-availability architecture provides two electrical paths to critical equipment.

Conceptually:

Path A

Utility/Generator A → Transformer A → LV Switchboard A → UPS A → Critical Distribution A

Path B

Utility/Generator B → Transformer B → LV Switchboard B → UPS B → Critical Distribution B

Dual-corded IT equipment can then receive:

Supply A + Supply B

If one path is unavailable, the second path continues supplying the equipment.

This architecture creates several important panel-design requirements.

The A and B paths may need:

• Physical segregation
• Independent busbars
• Separate incomers
• Separate protection
• Independent controls
• Separate auxiliary supplies
• Careful bus-coupler philosophy
• Clear identification
• Independent monitoring

The objective is to prevent a common failure from removing both paths.

The Bus Coupler: Useful but Potentially Dangerous

Consider:

Transformer A → Incomer A → Bus A

Transformer B → Incomer B → Bus B

with a:

Bus Coupler between A and B

The bus coupler can provide operational flexibility.

But it also introduces an important engineering question:

Can both sources operate in parallel?

If YES:

The system fault level may increase significantly.

Protection and synchronization requirements may change.

Transformer impedance and vector-group compatibility must be considered.

If NO:

The electrical and/or mechanical interlocking philosophy must prevent unintended parallel operation.

The RFQ should clearly state:

• Normal bus-coupler position
• Automatic/manual operation
• Source-paralleling permission
• Dead-bus logic
• Interlocking
• Transfer sequence
• Failure response
• Maintenance operating mode

Do not leave this logic for the panel manufacturer to guess.

This is the item we push back on at clarification stage, because it is the one clause that changes both the interlock scheme and the fault rating of the whole switchboard. If two transformers are permitted to run in parallel even briefly through a closed coupler, the prospective fault current at the bus is the combined contribution of both sources, not one. On two 1000 kVA transformers at 5% impedance, closing the coupler roughly doubles the transformer contribution at the busbar compared with either transformer alone — so a board specified for the single-transformer case can end up under-rated for a configuration the operators are allowed to create. The actual figure must come from the project short-circuit study, but the principle decides the enquiry. Our practice is to get the answer in writing before the GA is frozen, because retro-fitting a higher Icw busbar system into an approved footprint is not a small change.

PCC Panel Design for a Data Center

The main PCC or LV switchboard is one of the most important distribution points.

Its specification should clearly define:

• Rated operational voltage
• Rated current
• Busbar rating
• Neutral rating
• Short-circuit withstand
• Incomer configuration
• Bus coupler
• Breaker type
• Form of separation
• IP rating
• Metering
• Communication
• Protection
• Cable/busduct termination
• Future expansion

IEC 61439-1:2020 provides the general requirements for LV assemblies, while IEC 61439-2:2020 applies specifically to power switchgear and controlgear assemblies up to its defined voltage limits.

For Indian projects, BIS also identifies IS/IEC 61439 Part 2:2020 for low-voltage power switchgear and controlgear assemblies.

Short-Circuit Rating Must Be Based on the Actual Architecture

Data centers often contain multiple transformers, generators, UPS systems and possible bus ties.

That means fault level should never be selected from a generic statement such as:

“Use 50 kA panel.”

The required short-circuit withstand should be established from the project short-circuit study.

Consider:

• Transformer rating
• Transformer impedance
• Utility contribution
• Parallel transformers
• Generator contribution
• Bus-coupler condition
• Cable/busduct impedance
• Motor contribution where relevant
• Future source additions

Particularly important:

Normal Operation May Not Be the Worst Fault Case

Suppose Bus A and Bus B normally operate separated.

If the system permits temporary paralleling during transfer or maintenance, that configuration may produce a different fault level.

The panel and switchgear must be evaluated for the permitted operating configurations.

Icu, Ics and Icw Should Not Be Confused

Data-center specifications frequently contain several short-circuit terms.

For circuit breakers you may encounter:

• Icu — Ultimate short-circuit breaking capacity
• Ics — Service short-circuit breaking capacity
• Icw — Short-time withstand current

These values serve different purposes.

For selective protection architectures, short-time withstand capability can become especially important because upstream breakers may intentionally remain closed for a defined time while a downstream breaker clears the fault.

Therefore, breaker selection should consider the complete protection-coordination philosophy rather than simply selecting a breaker with a high headline kA rating.

Selectivity Is Critical in a Data Center

Imagine a short circuit on one downstream rack-distribution feeder.

The desirable result is:

Only that feeder trips.

Not:

The entire UPS output panel trips.

And certainly not:

The main data-center incomer trips.

This is the purpose of protection selectivity or discrimination.

The protection study should coordinate:

• Main incomer
• Bus coupler
• UPS input breakers
• UPS output breakers
• Distribution-board incomers
• PDU feeders
• Downstream breakers

Consider:

• Long-time settings
• Short-time settings
• Instantaneous settings
• Ground-fault settings
• Breaker selectivity tables
• Zone-selective interlocking where applicable
• Equipment withstand limits

Data-center switchgear should be designed around the protection study—not have settings decided only after the panel has been manufactured.

Form 4 Panels: Why Are They Commonly Considered?

Internal separation is often important in critical distribution systems.

Depending on the selected arrangement, IEC 61439-2 provides forms of internal separation for power switchgear assemblies.

Higher forms of separation can provide greater segregation between:

• Busbars
• Functional units
• Terminals

This can support maintenance and reduce the likelihood that work in one compartment exposes personnel or adjacent circuits to other live parts.

But there is an important engineering principle:

Higher Form Does Not Automatically Mean Higher Data-Center Tier

Form of separation is a panel construction feature.

Tier classification is a system topology and availability concept.

They support different objectives.

A Form 4 panel can be very valuable in a data-center architecture, but it does not replace redundant power paths, maintainable infrastructure or fault-tolerant topology.

Fixed vs Draw-Out Breakers

Critical main LV switchboards frequently use draw-out ACBs because they can facilitate:

• Isolation
• Maintenance
• Replacement
• Test position
• Reduced disturbance to fixed busbar connections

However, simply specifying every breaker as draw-out can dramatically increase:

• Panel dimensions
• Cost
• Maintenance requirements

A sensible philosophy might use:

Draw-out ACBs for critical incomers and bus couplers.

And suitable fixed or withdrawable MCCBs for downstream feeders depending on the operational requirement.

Again, maintainability should determine the construction—not marketing language.

ATS: Automatic Transfer Requires More Than Two Breakers

An ATS system transfers load between two power sources.

Typical sources could include:

Utility → Generator

or

Source A → Source B

A data-center ATS panel should clearly define:

• Preferred source
• Alternate source
• Voltage limits
• Frequency limits
• Transfer delay
• Return delay
• Re-transfer philosophy
• Open transition / other transfer method
• Dead-bus confirmation
• Generator start command
• Failure-to-transfer alarm
• Manual mode
• Test mode
• Interlocks
• Bypass requirement

The control system must also consider:

What happens when the preferred source becomes healthy again?

Immediate re-transfer may not always be desirable.

ATS vs STS

These are not the same device.

ATS — Automatic Transfer Switch

Generally uses electromechanical switching and transfers between available power sources according to a defined control sequence.

STS — Static Transfer Switch

Uses power electronics to perform very rapid transfer between sources for suitable critical-load applications.

For certain highly sensitive single-cord loads, STS systems may be considered downstream.

The practical differences are set out below.

Characteristic	ATS — Automatic Transfer Switch	STS — Static Transfer Switch
Switching element	Electromechanical contacts, contactors or breakers	Power semiconductors, commonly thyristors
Transfer behaviour	Transfers through a defined interruption, suited to loads that tolerate a short break	Very rapid transfer, intended for loads that cannot tolerate a break
Usual position in the power chain	Upstream — utility to generator, or source A to source B	Downstream — feeding sensitive critical loads
Source synchronisation	Not necessarily required; open-transition schemes transfer through a dead interval	Both sources generally need to be within defined limits for fast transfer
Best suited to	Cooling plant, essential and non-essential auxiliaries, general distribution	Single-corded IT equipment with no second power supply
Heat inside the panel	Negligible once the contacts are closed	Semiconductors dissipate continuously, so enclosure cooling must allow for it
Main selection driver	Source availability and the required transfer sequence	Load sensitivity and the permitted interruption time

The selection should depend on:

• Load sensitivity
• Source synchronization
• Transfer-time requirement
• Fault behaviour
• System architecture

Do not specify an STS simply because it is faster.

Our recommendation: deal with dual-corded equipment at the IT procurement stage rather than by adding static transfer switches later. An STS is genuinely necessary for a load that has only one power inlet, but every STS added downstream is a new series component in the critical path with its own failure modes and its own continuous heat load. Where the racks are properly dual-corded onto independent A and B paths, that redundancy is already achieved at the load itself, and the STS becomes an extra thing to maintain rather than a reliability gain.

3-Pole or 4-Pole ATS?

This is another commonly misunderstood question.

The answer should not simply be:

“Data centers always require 4-pole ATS.”

Whether the neutral must be switched depends on factors including:

• Earthing arrangement
• Generator neutral treatment
• Separately derived source philosophy
• Ground-fault protection
• Parallel neutral paths
• Project standards
• Transfer scheme

In some architectures, switching the neutral is essential.

In others, a solid neutral may be appropriate.

The decision must be coordinated with the system earthing study and source configuration.

Neutral Busbar Design Is Particularly Important

Data centers contain a large population of nonlinear electronic loads.

These may include:

• UPS systems
• Server power supplies
• IT equipment
• Switch-mode power supplies
• Lighting and control electronics

Therefore, neutral design should not be treated casually.

The engineer should consider:

• Harmonic spectrum
• Triplen harmonics
• Phase loading
• UPS topology
• Downstream distribution
• Neutral conductor sizing
• Thermal performance

A reduced neutral should not automatically be specified simply because a conventional balanced three-phase load would theoretically require less neutral current.

Actual system behaviour should determine the design.

UPS Input and Output Affect Panel Design Differently

A UPS sits between two different electrical environments.

UPS Input Side

The panel may need to consider:

• Rectifier input behaviour
• Harmonics
• Input power factor
• Bypass supply
• Battery charging
• Fault contribution
• Generator compatibility

UPS Output Side

The distribution system may need to consider:

• UPS inverter fault capability
• Static bypass
• Downstream breaker coordination
• Neutral loading
• Selectivity
• Transfer behaviour

A conventional utility transformer can often deliver high short-circuit current.

A UPS inverter may limit fault current very differently.

This is particularly important when coordinating downstream protection.

Do not assume the same protection settings will behave identically on utility bypass and UPS inverter operation.

Harmonics and Data-Center Panels

Modern power-electronic equipment can create nonlinear current.

Depending on the architecture, sources may include:

• UPS rectifiers
• VFDs for cooling systems
• Server power supplies
• Chargers
• Other electronic loads

The designer should evaluate:

• THDi
• THDv
• Transformer heating
• Neutral currents
• Capacitor interaction
• Generator operation
• Harmonic filter requirements

Avoid adding a generic harmonic filter to every data-center panel.

First establish:

Where is the harmonic problem?

What equipment produces it?

At which operating condition?

At what point must compliance be achieved?

Data Center Cooling Loads Also Matter

Not all data-center electrical loads are IT loads.

Significant power may also be required by:

• Chillers
• CRAH/CRAC systems
• Pumps
• Cooling towers
• Fans
• AHUs

Many of these motors may use VFDs.

Therefore, data-center electrical distribution also needs to consider:

• Motor starting
• VFD harmonics
• Generator loading
• Emergency cooling
• Load shedding
• Sequential restart
• Essential/non-essential classification

If utility power fails, starting every cooling motor simultaneously when the DG comes online may not be desirable.

The control philosophy may require staged restoration.

Generator Synchronization and Data Centers

Larger data centers may use multiple DG sets.

Instead of one generator feeding one load, the architecture may include:

Multiple DGs → Synchronization Panel → Emergency Bus

The synchronization/control system may need to manage:

• Automatic start
• Voltage matching
• Frequency matching
• Phase sequence
• Synchronizing
• Load sharing
• kW sharing
• kVAr sharing
• Generator sequencing
• Minimum running sets
• Load-based start/stop
• Failure handling
• Black-bus operation
• Load shedding

The panel manufacturer needs the complete generator operating philosophy.

Simply stating:

“Provide DG synchronization panel for 4 generators”

is not enough.

Metering and Monitoring Should Be Designed for Operations

Data centers need visibility into their electrical system.

Useful parameters may include:

• Voltage
• Current
• Frequency
• kW
• kVA
• kVAr
• Power factor
• Energy
• Demand
• Breaker status
• Trip status
• Temperature
• THD
• Generator status
• UPS status

Communication may be required to:

• BMS
• SCADA
• EPMS
• DCIM

An effective monitoring strategy should answer:

Where is energy being consumed?

How much capacity remains?

Which path is loaded?

What happened immediately before a trip?

Is redundancy still available?

Meters should therefore be selected as part of the operational strategy rather than simply installed because the specification says “multifunction meter.”

EPMS and DCIM Integration

Electrical Power Monitoring Systems can provide detailed visibility into the electrical infrastructure.

A data-center panel may therefore need:

• Intelligent trip units
• Multifunction meters
• Digital communication
• Breaker status
• Trip cause
• Energy data
• Alarm data
• Temperature-monitoring interfaces

Before manufacturing, define:

• Communication protocol
• Device addressing
• Network architecture
• Required data points
• Integration responsibility
• Cybersecurity/interface requirements

Otherwise the panel may be physically complete but difficult to integrate with the facility monitoring system.

Maintenance Is a Design Condition

One of the strongest questions to ask during data-center panel design is:

“How will this equipment be maintained after the facility becomes operational?”

Consider:

• Can the incomer breaker be withdrawn?
• Can one section be isolated?
• Can maintenance occur without interrupting the critical load?
• Can the bus coupler be serviced?
• Can an auxiliary supply be isolated?
• Can a failed meter be replaced safely?
• Is there sufficient access?
• Are terminals segregated?
• Is maintenance bypass available where needed?

ISO/IEC 22237 considers availability as a fundamental design characteristic across data-center infrastructure, while Uptime Tier III specifically focuses on concurrent maintainability of the infrastructure topology.

The panel should support the intended maintenance philosophy.

Common-Mode Failures Should Be Actively Considered

Redundancy loses value when both redundant systems depend on the same hidden component.

For example:

Two incomers may exist, but both rely on:

• One control transformer
• One PLC
• One DC supply
• One auxiliary MCB
• One communication network
• One common busbar section

That single component may become a common point of failure.

Ask during design:

If this component fails, can both A and B paths be affected?

True redundancy requires more than duplicating large equipment.

Auxiliary circuits matter too.

In our FAT experience, this is where a 2N drawing quietly turns into something less than 2N. The large equipment is nearly always properly duplicated, because it is visible on the single line diagram and it is expensive. The control circuits are not on the SLD, and that is exactly where a single control transformer, one 24 V DC supply, or one auxiliary MCB feeding both sections tends to appear. Our standard check during FAT is to switch off the control supply of section A with both sections energised and confirm that section B's breakers, protection, metering and status signals are entirely unaffected — then repeat it the other way round. We have seen designs where losing one auxiliary MCB removed the closing supply for both incomers and the bus coupler, so the switchboard could not be operated at all even though every power component was duplicated. Our recommendation: ask for independent control supplies per section, with a documented test in the ITP proving each section survives the loss of the other's auxiliaries.

Panel Thermal Design Is Becoming More Important

Higher current density can create significant thermal stress inside switchboards.

Panel designers should consider:

• Continuous current
• Busbar arrangement
• Breaker heat losses
• Harmonic loading
• Ambient temperature
• Ventilation
• Section arrangement
• Cable heat
• Adjacent heat-producing equipment

Do not assume:

Rated breaker current = panel continuous-current capability under every enclosure condition.

The complete assembly design and applicable verification must support the intended operating conditions.

Cable or Busduct?

Large data centers may use busduct/busway for high-current connections.

The decision between cable and busduct affects the panel interface.

For cable termination, consider:

• Number of runs
• Cable size
• Bending radius
• Gland space
• Lug arrangement

For busduct, consider:

• Connection orientation
• Phase sequence
• Neutral
• Earth
• Mechanical support
• Short-circuit forces
• Interface dimensions
• Future removal/access

The panel manufacturer should receive the busduct or cable interface requirement before finalizing the GA.

Future Expansion Should Be Engineered

Data-center load often grows in phases.

Therefore, define future requirements such as:

• Spare breakers
• Future spaces
• Busbar capacity
• Busbar extension
• Additional UPS feeders
• Additional cooling feeders
• Metering provisions
• Communication ports

But again:

“20% spare” is not enough.

Does 20% mean:

• 20% spare feeders installed?
• 20% blank space?
• 20% additional busbar current?
• 20% physical footprint?

Define it.

Data Center Electrical Panel Specification Checklist

Before sending an RFQ, define:

System

✓ Utility sources ✓ Transformer ratings ✓ Transformer impedance ✓ DG architecture ✓ UPS architecture ✓ A/B power paths ✓ Earthing arrangement ✓ Fault level

Availability

✓ N / N+1 / 2N philosophy ✓ Tier objective if applicable ✓ Concurrent maintenance requirement ✓ Fault-tolerance requirement ✓ Common-mode failure considerations

Main Switchboard

✓ Rated current ✓ Busbar rating ✓ Neutral rating ✓ Short-circuit withstand ✓ Form of separation ✓ IP rating ✓ Draw-out/fixed breaker philosophy

Source Transfer

✓ ATS arrangement ✓ Preferred source ✓ Alternate source ✓ 3P/4P requirement ✓ Bus-coupler logic ✓ Source paralleling permitted/not permitted ✓ Transfer delays ✓ Return philosophy

Protection

✓ Protection study ✓ Selectivity ✓ ACB settings ✓ Earth-fault philosophy ✓ Zone-selective interlocking if required

Monitoring

✓ Metering points ✓ Energy meters ✓ THD monitoring ✓ Communication protocol ✓ EPMS/BMS/DCIM interface

Installation

✓ Cable entry ✓ Busduct interface ✓ Cable alley ✓ Termination space ✓ Panel access ✓ Maintainability

Quality

✓ IEC/IS compliance ✓ Design-verification evidence ✓ FAT procedure ✓ Customer/TPI witness requirements ✓ Routine-verification documentation ✓ Calibration records

Common Data Center Panel Design Mistakes

Mistake 1: Calling a Panel “Tier III” or “Tier IV”

Tier classification applies to the infrastructure topology, not an individual switchboard.

Mistake 2: Assuming Form 4 Means Tier IV

Form of separation and availability topology are different engineering concepts.

Mistake 3: Installing a Bus Coupler Without Defining Parallel Operation

This can affect fault level and protection coordination.

Mistake 4: Selecting 3-Pole or 4-Pole ATS Without Studying the Earthing System

Neutral switching should be based on the source and earthing philosophy.

Mistake 5: Designing Only for Normal Operation

Maintenance and failure conditions may define the real requirement.

Mistake 6: Ignoring UPS Fault Characteristics

Downstream protection must work under both inverter and bypass operating conditions.

Mistake 7: Using Reduced Neutral Without Evaluating Nonlinear Loads

Electronic loads can materially affect neutral loading.

Mistake 8: Creating Two Power Paths With Common Auxiliary Failure Points

Redundant main equipment does not guarantee redundant operation.

Mistake 9: Specifying the Highest Form, IP and Breaker Rating Without Engineering Need

Higher specification can increase cost and size without solving the actual availability problem.

Mistake 10: Ignoring Future Expansion

A panel that works perfectly on Day 1 may become difficult to expand after the data center is live.

How Wisdom Techno Solutions Approaches Data Center Panel Requirements

Data-center electrical distribution requires closer coordination between the consultant, system designer, EPC contractor, panel manufacturer, UPS supplier, DG supplier and automation team than a conventional LT panel project.

Wisdom Techno Solutions is a licensed partner for three design-verified switchgear platforms: Rittal Ri4Power (Rittal Solutions Partner since 2021 — the Ri4Power system carries Intertek ASTA design verification to IEC 61439-1 and -2 Edition 3, with copper and aluminium busbar systems up to 6300 A), C&S CX (Licensed Partner since 2023, up to Form 4B Type II) and Siemens SIEPAN Elite/8PU (IEC 61439 Solutions Partner since 2024). We build to each system's verified configuration and routine-verify every assembly in-house at our Vadodara, Gujarat facility, for industrial, EPC and infrastructure projects across India. That places the A/B-path busbar rating and form of separation inside a verified design envelope.

We engineer and manufacture customized power and control assemblies based on approved project requirements, including applications involving:

• Main LT PCC panels
• Power distribution panels
• MCC/PMCC panels for cooling and auxiliary systems
• ATS / AMF panels
• DG synchronization panels
• VFD panels
• PLC and automation panels
• Metering and communication integration

For a data-center application, WTS would begin with the system architecture rather than assuming a standard panel configuration.

Useful inputs include:

• SLD
• A/B distribution philosophy
• Transformer data
• Generator architecture
• UPS architecture
• Fault-level study
• Protection study
• Bus-coupler philosophy
• Form requirement
• Neutral/earthing philosophy
• Cable or busduct details
• Metering and communication requirements
• Approved makes
• FAT/inspection requirements

The objective should be to manufacture the panel around the availability and operating philosophy of the data center, rather than force the data center architecture into a standard panel BOM.

FAT for Data Center Panels

A data-center switchboard FAT should go significantly beyond appearance.

Depending on project scope, FAT should verify:

• Approved GA
• SLD
• BOM
• Busbar construction
• A/B section identification
• Breaker ratings
• Mechanical operation
• Bus-coupler interlocks
• ATS logic
• Generator commands
• Local/remote operation
• Protection trips
• Metering
• Communication
• PLC logic
• Alarm sequences
• Fail-safe behaviour
• Auxiliary-supply failure
• Applicable HV/IR testing
• Protective continuity
• Test documentation

Particularly important:

Test Failure Scenarios

Do not test only the happy path.

Simulate:

• Source A failure
• Source B failure
• Breaker failure
• ATS failure alarm
• Bus-coupler prohibited condition
• Control-power loss
• Communication loss
• Emergency trip

A critical-power panel should demonstrate how it behaves when something goes wrong.

WTS In-House Testing and TPI Capability

Wisdom Techno Solutions performs applicable panel FAT and electrical testing in-house using calibrated testing instruments and equipment.

Depending on the project, testing can include:

• HV / dielectric testing
• Insulation resistance testing
• Busbar/current-path related checks
• Protective continuity
• Wiring verification
• Breaker operation
• Interlocks
• Protection functions
• Metering
• PLC/HMI functions
• Communication testing

Customer and consultant representatives can witness FAT where required.

Where specified by the project, we also facilitate Third Party Inspection (TPI) according to the agreed Inspection and Test Plan.

For a data-center project, this testing discipline becomes especially important because a small control or interlock mistake can affect the availability architecture of the complete facility.

Applicable Standards and Guidance

Depending on project location and specification, relevant references may include:

IEC 61439-1:2020

General rules for low-voltage switchgear and controlgear assemblies.

IEC 61439-2:2020

Specific requirements for power switchgear and controlgear assemblies.

IS/IEC 61439 Part 2:2020

Indian adoption applicable to power switchgear and controlgear assemblies.

ISO/IEC 22237-1:2021

General concepts for data-center facilities and infrastructures.

ISO/IEC 22237-3:2021

Specific data-center infrastructure standard covering power distribution.

Uptime Institute Tier Standard

Provides site-infrastructure topology criteria based on increasing levels of redundant capacity and distribution paths.

The final project specification should identify the applicable standards, local regulations and client requirements rather than using the generic phrase:

“Panel shall comply with all international standards.”

Conclusion

Reliable data-center power distribution does not come from simply buying higher-rated switchgear.

It comes from designing the complete electrical architecture around:

Availability

Redundancy

Maintainability

Fault containment

Protection selectivity

and

Visibility

Start with the power topology.

Define A and B paths.

Understand N, N+1 and 2N.

Establish the maintenance philosophy.

Calculate the fault level.

Coordinate protection.

Decide the neutral and earthing arrangement correctly.

Engineer ATS and bus-coupler logic.

Consider UPS fault behaviour.

Account for harmonic loads.

Provide useful monitoring.

Remove common points of failure.

Then design the panel.

A well-engineered data-center panel should not simply distribute electricity.

It should help ensure that when one part of the electrical system is maintained or fails, the critical load continues to receive power according to the intended facility architecture.

Planning a data-center PCC, ATS, synchronization, VFD or critical power-distribution panel?

Share the SLD, redundancy philosophy, fault-level data, UPS/DG architecture and technical specification with Wisdom Techno Solutions for project-specific panel engineering and manufacturing.

Related Guides

• 3-pole vs 4-pole ATS
• Form 1/2/3/4a/4b internal separation
• Internal arc and IEC TR 61641
• BESS electrical panel design
• Product page: IEC 61439 type-tested panels

Frequently Asked Questions

What electrical panels are used in a data center?

Depending on the architecture, data centers may use main PCC/LV switchboards, ATS panels, generator synchronization panels, UPS input/output distribution panels, PDUs, MCCs for cooling systems, VFD panels and downstream critical distribution boards. These fall into two broad families: the critical path that feeds the IT load through UPS systems, and the mechanical path that feeds chillers, CRAH units, pumps and cooling towers. Both are essential, because a data center that loses cooling will shut down almost as quickly as one that loses IT power. The panel schedule therefore follows from the facility's availability architecture rather than from a standard list of boards.

What is N+1 redundancy in a data center?

N+1 means the system has the capacity required for the load plus one additional redundant capacity component. If the critical load needs three UPS modules, then three required plus one redundant module is N+1, and any one module can be taken out of service while sufficient capacity remains. The important limitation is that N+1 describes capacity components only — it says nothing about the number of distribution paths to the load. A facility can be N+1 on UPS capacity and still lose the critical load from a single fault in its one distribution path.

What is 2N redundancy?

2N generally provides two independent capacity systems or paths, each capable of supporting the required critical load on its own. Dual-corded IT equipment then takes an A supply and a B supply, so an entire path can be isolated for maintenance or lost to a fault without interrupting the load. The word doing the real work in that definition is "independent" — if both paths share a transformer, a control supply, a PLC or a busbar section, the architecture is not truly 2N regardless of what the drawing is labelled. Auxiliary and control circuits must be duplicated along with the power equipment.

What is the difference between Tier III and Tier IV?

Uptime Institute describes Tier III infrastructure as concurrently maintainable and Tier IV as fault tolerant, with different requirements for redundant components and distribution paths. Concurrent maintainability means any capacity component or distribution element can be removed from service on a planned basis without affecting the critical load. Fault tolerance is the stronger requirement: the facility must also withstand an unplanned failure of a component or path without dropping the load. Both are classifications of the complete site infrastructure topology, so neither can be met or claimed by the specification of an individual switchboard.

Is Form 4 required for a Tier IV data center?

A Tier classification cannot be determined from the panel Form alone. Form of internal separation is a switchboard construction characteristic, while Tier classification evaluates the site infrastructure topology.

Does a Form 4 panel make a data center Tier IV?

No. Form 4 and Tier IV refer to completely different concepts.

Why are draw-out ACBs commonly used in critical switchboards?

Draw-out construction can facilitate isolation, testing, maintenance and replacement of critical breakers without disturbing fixed busbar connections, depending on the architecture. The test position also allows the breaker's closing, tripping and protection functions to be exercised with the main contacts disconnected from the power circuit, which supports maintenance while the board remains live. That said, specifying every breaker as draw-out significantly increases panel dimensions, cost and the maintenance regime itself. A common approach is draw-out ACBs for critical incomers and bus couplers, with fixed or withdrawable MCCBs on downstream feeders where the operational requirement does not justify draw-out.

Should a data-center ATS be 3-pole or 4-pole?

There is no universal answer. Neutral switching should be determined from the earthing arrangement, source configuration, generator neutral philosophy and protection design.

Why is neutral sizing important in data centers?

Data centers contain significant nonlinear electronic loads. Neutral current and harmonic behaviour should therefore be considered instead of automatically selecting a reduced neutral.

Why is protection selectivity important?

Selectivity aims to isolate the fault using the nearest downstream protective device rather than unnecessarily disconnecting a larger part of the data center. A short circuit on one rack distribution feeder should trip only that feeder, not the UPS output panel and certainly not the main incomer. Achieving this requires a protection study that coordinates long-time, short-time, instantaneous and earth-fault settings across the whole chain, and it is why short-time withstand capability matters — an upstream breaker may need to hold a fault current for a defined interval while the downstream device clears it. Settings should be established from the study before manufacture, not decided after the panel is built.

What is the difference between ATS and STS?

An ATS typically transfers between sources through electromechanical switching, while an STS uses power electronics for very rapid transfer in suitable critical-load applications. An ATS normally sits upstream, transferring between utility and generator or between source A and source B, and it transfers through a defined interruption that general and mechanical loads can tolerate. An STS sits downstream and is intended for sensitive loads that cannot tolerate any break — typically single-corded IT equipment with no second power inlet. An STS also dissipates heat continuously through its semiconductors and adds a component in the critical path, so it should be selected for a genuine load-sensitivity requirement rather than because it is faster.

Which standard specifically covers data-center power distribution?

ISO/IEC 22237-3:2021 is titled Data centre facilities and infrastructures — Part 3: Power distribution, and it addresses the power distribution infrastructure of a data center specifically. It forms part of the wider ISO/IEC 22237 series, whose Part 1 (ISO/IEC 22237-1:2021) covers general concepts for data-center facilities and infrastructures. This is separate from the standards governing the equipment itself: an LV switchboard supplied into a data center is assessed against IEC 61439-1 and IEC 61439-2. The Uptime Institute Tier Standard is a different kind of document again — it provides site-infrastructure topology criteria rather than product or installation requirements.

Which IEC standard applies to the main LT switchboard?

IEC 61439-1:2020 gives the general rules for low-voltage switchgear and controlgear assemblies, and IEC 61439-2:2020 gives the specific requirements for power switchgear and controlgear assemblies — the main LT switchboard falls under both. For Indian projects, BIS identifies IS/IEC 61439 Part 2:2020 for the same class of assembly. These standards cover the assembly as a product, including its design verification and routine verification requirements and its forms of internal separation. They do not address facility availability or Tier classification, which are governed by entirely separate documents.

Can Wisdom Techno Solutions perform customer-witnessed FAT?

Yes. We perform applicable in-house testing using calibrated equipment, support customer/consultant-witnessed FAT and facilitate TPI where required by the project.

What information should be sent for a data-center panel quotation?

Ideally provide the SLD, transformer and DG data, UPS architecture, A/B distribution philosophy, fault level, protection requirements, bus-coupler/ATS logic, neutral and earthing philosophy, Form/IP requirements, cable or busduct details, metering/communication requirements and FAT specification. The two inputs that most often decide the panel's size and cost are the short-circuit study result and whether source paralleling through the bus coupler is permitted, so state both explicitly rather than leaving them to be assumed. Also confirm the earthing arrangement and generator neutral treatment early, because those determine whether the ATS needs to switch the neutral. If the protection study is not yet complete, say so at enquiry stage — the busbar and breaker selection depend on it, and changing them after the general arrangement drawing is approved is expensive.